JP4237259B2 - Chip package substrate and manufacturing method thereof - Google Patents

Description

例１
微細分散液は、難燃化したジシアナミド／２−メチルイミダゾールに触媒されるビスフェノール−Ａに基づくポリグリシジルエーテル（Ｎｅｌｃｏ Ｎ−４００２−５，Ｎｅｌｃｏ Ｃｏｒｐ）の２０％（ｗ／ｗ）ＭＥＫ溶液に、２８１．６ｇのＴｉＯ₂（ＴＩ Ｐｕｒｅ Ｒ−９００，Ｄｕ Ｐｏｎｔ Ｃｏｍｐａｎｙ）を混合して調整した。この分散液は、均一性を確実にするために、絶えず撹拌した。その後、延伸膨張ＰＴＦＥの布きれをこの樹脂混合物に浸した。柔軟な複合材料を与えるために、ウェブを引っ張り応力をかけて１６５℃で１分間乾燥した。この様に作られた部分的に硬化した接着剤複合材料は、５７重量％のＴｉＯ₂、１３重量％のＰＴＦＥおよび３０重量％のエポキシ接着剤を含んでいた。接着剤シートのいくつかのプライを銅箔の間に重ねて、２２５℃の温度で９０分間、真空補助油圧プレスを使用して６００ｐｓｉでプレスし、その後加圧下で冷却した。この結果、誘電率が１９．０で、平均の層の厚さが１００μｍ（０．００３９”（３．９ｍｉｌ））の誘電体積層体厚さで２８０℃のはんだ付けの衝撃に３０秒間耐える銅の積層板ができた。
例２
微細分散液は、フェニルトリメトキシシラン（０４３３０，Ｈｕｌｓ／Ｐｅｔｒａｒｃｈ）で前処理した３８６ｇのＳｉＯ₂（ＨＷ−１１−８９，Ｈａｒｂｉｓｏｎ Ｗａｌｋｅｒ Ｃｏｒｐ．）を２００ｇのビスマレイミドトリアジン樹脂（ＢＴ２０６ＯＢＪ，三菱ガス化学）および３８８ｇのＭＥＫのマンガン触媒入り溶液に混ぜ入れて調整した。この分散液は、均一性を確実にするために、絶えず撹拌した。その後、０．０００２”の厚さの延伸膨張ＰＴＦＥの布きれをこの樹脂混合物に浸して、取り出し、そしてその後で柔軟な複合材料を与えるために、引っ張り応力をかけて１６５℃で１分間乾燥した。このプレプレグのいくつかのプライを銅箔の間に重ねて、２２５℃の温度で９０分間、真空補助油圧プレスを使用して２５０ｐｓｉでプレスし、その後加圧下で冷却した。結果としてこの様に作られた誘電体は、５３重量％のＳｉＯ₂、５重量％のＰＴＦＥおよび４２重量％の接着剤を含んでおり、銅に良く密着して、誘電率が（１０ＧＨｚにおいて）３．３および誘電正接が（１０ＧＨｚにおいて）０．００５であることを示した。
例３
微細分散液は、４８３ｇのＳｉＯ₂（ＨＷ−１１−８９）を、２７４．７ｇのビスマレイミドトリアジン樹脂（ＢＴ２０６０ＢＪ，三菱ガス化学）および４８５ｇのＭＥＫのマンガン触媒入り溶液に混ぜ入れて調整した。この分散液は、均一性を確実にするために、絶えず撹拌した。その後、０．０００２”の厚さの延伸膨張ＰＴＦＥの布きれをこの樹脂混合物に浸して、取り出し、そしてその後で柔軟な複合材料を与えるために、引っ張り応力をかけて１６５℃で１分間乾燥した。このプレプレグのいくつかのプライを銅箔の間に重ねて、２２５℃の温度で９０分間、真空補助油圧プレスを使用して２５０ｐｓｉでプレスし、その後加圧下で冷却した。結果としてこの様に作られた誘電体は、５７重量％のＳｉＯ₂、４重量％のＰＴＦＥおよび３９重量％の接着剤を含んでおり、銅に良く密着して、誘電率が（１０ＧＨｚにおいて）３．２および誘電正接が（１０ＧＨｚにおいて）０．００５であることを示した。
例４
微細分散液は、１５．４４ｋｇのＴｉＯ₂粉末（ＴＩ Ｐｕｒｅ Ｒ−９００，ＤｕＰｏｎｔ Ｃｏｍｐａｎｙ）を、３．３０ｋｇのビスマレイミドトリアジン樹脂（ＢＴ２０６ＯＢＨ，三菱ガス化学）および１５．３８ｋｇのＭＥＫマンガン触媒入り溶液に混ぜ入れて調整した。この分散液は、均一性を確実にするために、絶えず撹拌した。その後０．０００４”のＴｉＯ₂入りの延伸膨張ＰＴＦＥの布きれ（Ｍｏｒｔｉｍｅｒの米国特許第４，９８５，２９６号明細書の手法により充填した、ただしＴｉＯ₂添加量は４０％とし、最後に膜を圧縮しなかった）を、樹脂混合物に浸して、取り出し、そしてその後で柔軟な複合材料を与えるために、引っ張り応力をかけて１６５℃で１分間乾燥した。このように作られた部分的に硬化した接着性複合材は、７０重量％のＴｉＯ₂、９重量％のＰＴＦＥおよび２１重量％の接着剤を含んでいた。このプレプレグのいくつかのプライを銅箔の間に重ねて、２２０℃の温度で９０分間、真空補助油圧プレスを使用して５００ｐｓｉでプレスし、その後加圧下で冷却した。このように得られた誘電体は、銅に良く密着して、誘電率が１０．０、および誘電正接が０．００８であることを示した。
例５
微細分散液は、７．３５ｋｇのＳｉＯ₂（ＡＤＭＡＴＥＣＨＳ ＳＯ−Ｅ２，タツモリ社）を、７．３５ｋｇのＭＥＫおよび７３．５ｇのカップリング剤、すなわち３−グリシジルオキシプロピルトリメトキシシラン（Ｄｙｎａｓｙｌａｎ ＧＬＹＭＯ（Ｐｅｔｒａｃｈ Ｓｙｓｔｅｍ））と混合して調整した。ＳＯ−Ｅ２は、粒子サイズが０．４〜０．６ｍｍ、比表面積が４〜８ｍ²／ｇ、および嵩密度が０．２〜０．４ｇ／ｃｃ（ゆるい（ｌｏｏｓｅ））の非常に球状なシリカを持つものと製造者により説明されている。
この分散系に、シアノ化したフェノール樹脂、Ｐｒｉｍａｓｅｔ ＰＴ−３０（Ｌｏｎｚａ Ｃｏｒｐ．）の５０％（ｗ／ｗ）メチルエチルケトン（ＭＥＫ）溶液９３２ｇ、ＲＳＬ １４６２（Ｓｈｅｌｌ Ｒｅｓｉｎｓ，Ｉｎｃ．（ＣＡＳ＃２５０６８−３８−６））の５０％（ｗ／ｗ）ＭＥＫ溶液８９６ｇ、ＢＣ−５８（Ｇｒｅａｔ Ｌａｋｅｓ，Ｉｎｃ．）の５０％（ｗ／ｗ）ＭＥＫ溶液３８０ｇ、ビスフェノールＡ（Ａｌｄｒｉｃｈ Ｃｏｍｐａｎｙ）の５０％ＭＥＫ溶液５４ｇ、Ｉｒｇａｎｏｘ １０１０（Ｃｉｂａ Ｇｅｉｇｙ）１２．６ｇ，マンガン２−エチルヘキサノエート（Ｍｎ−ＨＥＸ−ＣＥＭ（ＯＭＧ Ｌｔｄ．））の０．６％溶液３．１ｇ、および２．４０ｋｇのＭＥＫを加えた。この分散系を約１〜３ガロン／分の速度で約２０分間Ｍｉｓｏｎｉｃｓ連続フローセルに通して超音波撹拌にかけた。このように得られた微細分散系を、浴全体の固形分濃度が１１．９％（ｗ／ｗ）になるように、更に薄めた。
微細分散系を含浸浴に注いだ。延伸膨張ポリテトラフルオロエチレンウェブは図１０のノードフィブリル構造物を持ち、以下のような物性を持つ。

フレーザー（Ｆｒａｚｉｅｒ）数は試験される物質の通気性に関係する。通気性は、空気流量を計るためのおよそ６平方インチの円形の領域に設けられるガスケット付きの掴み具にウェブを固定することで測定する。上流側を乾燥した圧縮空気の供給源を持つ管路内の流量計につないだ。試料掴み具の下流側は大気に解放した。試験は、０．５インチ水柱の圧力を試料の上流側にかけ、インライン流量計（流量計につないだボールフロートロータメータ）を通る空気の流量を記録して行う。
ボール破裂強さは破壊時の最大値を測定することにより、試料の相対強さを計る試験である。ウェブを２つの板の間に固定しながら、直径１インチのボールで試験する。ＣｈａｔｉｌｌｏｎのＦｏｒｃｅ Ｇａｕｇｅ Ｂａｌｌ／Ｂｒｕｓｔ Ｔｅｓｔを使用した。媒体を計測装置に緊張させて配置し、ウェブを破裂プローブのボールに接触させるように持ち上げることで圧力をかける。破壊時の圧力を記録する。
上で記述したウェブを、均一性を確実にするために約３フィート／分の速度で、常に撹拌している含浸浴に通過させた。含浸させたウェブを、全てまたはほとんど全ての溶媒を取り除くために、加熱した炉を直ぐに通して、ロールに集めた。
このプレプレグのいくつかのプライを銅箔の間に重ねて、２２０℃の温度で９０分間、真空補助油圧プレスを使用して２００ｐｓｉでプレスし、その後加圧下で冷却した。このように得られた誘電体は、銅に良く接着して、誘電率（１０ＧＨｚで）が３．０、および誘電正接が（１０ＧＨｚで）０．００８５であることを示した。
例４および例７で使用した微粒子充填剤の物理的性質を以下で比較する。

例６
融解ケイ素の蒸気燃焼から調整したＳｉＯ₂に基づく、含浸した接着剤充填剤混合物を含むｅＰＴＦＥマトリクッスを以下のように調整した。２つの先駆物質混合物を最初に調整した。１つはシリカが例５と同様のシラン処理したシリカを含むスラリーの形をしており、他方は樹脂および他の成分の触媒なしの配合物である。
混合物Ｉ
このシリカスラリーは、例５のＭＥＫ中のＳＯ−Ｅ２シリカの５０／５０配合物で、シリカはシリカの重量の１％に等しい被覆シランを含む。５ガロンの容器に１７．５ポンドのＭＥＫおよび７９ｇのシランを加え、二つの成分を混ぜてＭＥＫへのシランの均一な分散を確実にした。その後、例５の１７．５ポンドのシリカを加えた。５ガロンの容器２つ分のＭＥＫ−シリカ−シラン混合物を反応容器に加え、内容物、すなわちスラリーをおよそ１時間超音波分散器に通し、再循環させて、存在し得る全てのシリカ集合体を破壊した。超音波処理を完了し、その後反応器の内容物を連続的に撹拌しながら、この内容物をおよそ８０℃の温度でおよそ１時間加熱した。反応した混合物はその後１０ガロンの容器に移した。
混合物ＩＩ
所定の樹脂配合物製品はおよそ６０％の固形分を含む触媒なしの樹脂配合物（接着剤）を含有しているＭＥＫに基づく溶液であって、固形分はいずれも重量百分率で、ＰＴ−３０シアン化フェノール樹脂が４１．２％、ＲＳＬ １４６２エポキシ樹脂が３９．５％、ＢＣ５８難燃材が１６．７％、Ｉｒｇａｎｏｘ １０１０安定剤が１．５％、およびビスフェノールＡ助触媒が１％の厳密な混合物であった。
１０ガロンの容器の中に、１４．８ポンドのＰＴ−３０および１５〜２０ポンドのＭＥＫを加え、激しく撹拌して完全にＰＴ−３０を溶かした。その後６ポンドのＢＣ５８を計り、ＭＥＫ／ＰＴ−３０溶液に加えて、激しく撹拌し、ＢＣ５８を溶かした。安定剤、２４４．５ｇのＩｒｇａｎｏｘ １０１０およびビスフェノールＡ １６３ｇを加えた。１０ガロンの容器を再秤量し、１４．２２ポンドのＲＳＬ １４６２を加えた。更なるＭＥＫを加えて混合物の重さを６０ポンドにした。この内容物を１〜２時間、または固形成分が完全に解けるのに必要な時間激しく撹拌した。
所望の製品は、シランで処理したシリカ、触媒なしの樹脂配合物、およびＭＥＫの混合物であり、固形分の６８重量％はシリカ、および全固形分は混合物の５〜５０重量％である。厳密な固形分濃度は実験ごとに変わり、含浸させる膜にも一部依存する。触媒の濃度はＰＴ−３０およびＲＳＬ １４６２の合計の１０ｐｐｍであった。
混合物ＩおよびＩＩの固形分含有量を測定して、先駆物質の精度を検定し、それまでに起こった全ての溶媒のフラッシュを補償した。その後、混合物Ｉを１０ガロンの容器に加えて１２ポンドの固形分、例えば５１．５％が固形分なら２３．４８ポンドの混合物Ｉを供給した。次に混合物ＩＩを容器に加えて５．６４ポンドの固形分、例えば５９．６％が固形分なら９．４６ポンドの混合物ＩＩを供給した。３．４５ｇのマンガン触媒溶液（ミネラルスピリット中０．６％）を混合物ＩおよびＩＩの混合物に加え、徹底的に混合して高固形分含有量の混合物を作った。
ｅＰＴＦＥマトリクッスを含浸するための浴の混合物、すなわち２８％が固形分混合物であるこの高固形分含有量の混合物に十分なＭＥＫを加えて全体の重さを６３ポンドにすることで調整した。
その後、ｅＰＴＦＥマトリクッスをこの浴混合物で含浸して、誘電体物質を作った。
例７
微細分散液を、２６．８ｇのファーネスブラック（Ｓｐｅｃｉａｌ Ｓｃｈｗａｒｚ １００，Ｄｅｇｕｓｓａ Ｃｏｒｐ．，Ｒｉｄｇｅｆｉｅｌｄ Ｐａｒｋ，Ｎｅｗ Ｊｅｒｓｅｙ）および７９ｇのカップリング剤、３−グリシジルオキシプロピル−トリメトキシシラン（Ｐｅｔｒａｃｈ Ｓｙｓｔｅｍ）を混合して調整した。分散系を１分間超音波撹拌にかけ、その後で予め超音波撹拌された１７．５ポンドのＭＥＫ中で１７．５ポンドのＳｉＯ₂（ＳＯ−Ｅ２）が撹拌されている分散系に加えた。最終的な分散系を、還流下で１時間一定な上部混合（ｏｖｅｒｈｅａｄ ｍｉｘｉｎｇ）をしながら加熱し、その後室温まで冷却させた。
別に、接着剤ワニスを、５７．５％（ｗ／ｗ）Ｐｒｉｍａｓｅｔ ＰＴ−３０のＭＥＫ混合物３４１３ｇ、７６．８％（ｗ／ｗ）ＲＳＬ １４６２のＭＥＫ混合物２４５６ｇ、５３．２％（ｗ／ｗ）ＢＣ５８（Ｇｒｅａｔ ｌａｋｅｓ，Ｉｎｃ．）のＭＥＫ溶液１４９５ｇ，２３．９％（ｗ／ｗ）ビスフェノールＡ（Ａｌｄｒｉｃｈ Ｃｏｍｐａｎｙ）のＭＥＫ溶液２００ｇ、７１．５ｇのＩｒｇａｎｏｘ １０１０、０．６％（ｗ／ｗ）Ｍｕ ＨＥＸ−ＣＥＭ（ＯＭＧ Ｌｔｄ．）のミネラルスピリット溶液３．２１ｇ、および２．４０ｋｇのＭＥＫを加えて調整した。
別の容器に、上記分散系３７３９ｇを、ファーネスブラック（Ｓｐｅｃｉａｌ Ｓｃｈｗａｒｚ １００，Ｄｅｇｕｓｓａ Ｃｏｒｐ．，Ｒｉｄｇｅｆｉｅｌｄ Ｐａｒｋ，Ｎｅｗ Ｊｅｒｓｅｙ）０．０２３３ｇ、上記接着剤ワニス１３２８、およびＭＥＫ３８．３ポンドとともに入れた。この混合物を含浸浴に注ぎ、ｅＰＴＦＥウェブを３フィート／分または約３フィート／分の速度で含浸浴に通した。この分散系を常に撹拌して均一性を確実にした。含浸されたウェブをすぐに加熱した炉に通し、全てまたはほとんど全ての溶媒を除去し、ロールに集めた。
このプレプレグのいくつかのプライを銅箔の間に重ねて、２００℃の温度で９０分間、真空補助油圧プレスを使用して２００ｐｓｉでプレスし、その後加圧下で冷却した。結果として得られた誘電体は銅に対する良好な接着性を示した。
例８
接着剤ワニスを、５７．５％（ｗ／ｗ）Ｐｒｉｍａｓｅｔ ＰＴ−３０（Ｐ．Ｎ．Ｐ−８８−１５９１）のＭＥＫ溶液３４１３ｇ、７６．８％（ｗ／ｗ）ＲＳＬ １４６２のＭＥＫ溶液２４５６ｇ、５３．２％（ｗ／ｗ）ＢＣ５８（Ｇｒｅａｔ ｌａｋｅｓ，Ｉｎｃ．）のＭＥＫ溶液１４９５ｇ、２３．９％（ｗ／ｗ）ビスフェノールＡ（Ａｌｄｒｉｃｈ Ｃｏｍｐａｎｙ）のＭＥＫ溶液２００ｇ、Ｉｒｇａｎｏｘ １０１０ ７１．５ｇ、０．６％（ｗ／ｗ）Ｍｎ ＨＥＸ−ＣＥＭのミネラルスピリット溶液３．２１ｇ、およびＭＥＫ２．４０ｋｇを加えて調整した。
別の容器に、上記接着剤ワニス分散系１３２８ｇ、ＭＥＫ４２．３ポンド、ファーネスブラック（Ｓｐｅｃｉａｌ Ｓｃｈｗａｒｚ １００，Ｄｅｇｕｓｓａ Ｃｏｒｐ．，Ｒｉｄｇｅｆｉｅｌｄ Ｐａｒｋ，Ｎｅｗ Ｊｅｒｓｅｙ）６．４ｇおよび、ＳｉＯ₂（ＳＯ−Ｅ２）１８６０．９ｇを入れた。この混合物を含浸浴に注ぎ、ｅＰＴＦＥウェブを３フィート／分または約３フィート／分の速度で含浸浴に通した。この分散系を常に撹拌して均一性を確実にした。含浸されたウェブを直ぐに加熱した炉に通し、全てまたはほとんど全ての溶媒を除去し、ロールに集めた。
このプレプレグのいくつかのプライを錫箔の間に重ねて、２２０℃の温度で９０分間、真空補助油圧プレスを使用して２００ｐｓｉでプレスし、その後加圧下で冷却した。結果として得られた誘電体は銅に対する良好な接着性を示した。
上記のように、本発明はチップパッケージ用の寸法的に安定な芯を供給する。この芯は、例えば銅または銅でメッキしたＩＮＶＡＲのような金属芯である。銅はチップパッケージのＣＴＥを、それを結合するプリント回路板のＣＴＥにあわせやすいので、好ましい材料の一つである。このことは、金属がパッケージに必要な安定性を供給するので、パッケージを扱いやすくする。チップパッケージの上面と下面にメッキし、層を同時に積層することは、チップパッケージの製造工程をより効率的にするのを可能にする。加えて、基板のそれぞれの面のそれぞれの層の金属量を均衡させて、パッケージをチップに取り付けやすくする。金属芯の更なる利点には、チップの同時スイッチングノイズを減少させる埋め込み静電容量を提供することが含まれる。
図１１に示したように、参照してここに組み入れられる米国特許第５，４９８，４６７号明細書で教示されるような、伝導性のＺ軸材を使用する多層基板を使用してもよい。例えば、多層をこのような伝導性Ｚ軸材を使用して一緒に結合し、非常に密なプリント回路板が得られる。１００またはそれ以上の層に達すると考えることができよう。図１１では、基板１４０および１５０はそれぞれ互いに同一であり、構造的に図５の寸法的に安定な基板に類似する。基板１４０および１５０は、以下で詳細に説明される伝導性のＺ軸材１６０と共に積層される。図１９は、以下で図１２〜１４を参照に説明する伝導性のＺ軸材１６０と共に積層する図８の基板と類似の、基板２９０および３００を示す。
図１１で例示される多層パッケージは、Ｚ軸材の肉厚を通って延在する不規則な形の伝導経路を持つ上で挙げた米国特許第５，４９８，４６７号明細書に記載されるような、伝導性のＺ軸材を使用して作製する。伝導性Ｚ軸材を形成するための基材は、一つの面から他方の面への連続な気孔を持つ平板の、連続気泡の、多孔質部材である。この多孔質で平板の部材は、気孔を画定する材料が図１２〜１４で示すようにＺ軸平面を通る垂直に画成された断面内でＺ軸方向に通る不規則な経路を作る内部形態を持っていなければならない。
伝導性のＺ軸部材のふさわしい材料は、５×１０^-6ｍ〜５×１０^-4ｍ（５〜５００μｍ）程度の範囲の厚さを持ち、これにはナイロンといったような織布または不織布、グラスファイバーまたはポリエステル布帛または綿等が含まれる。この部材は、可撓性である多孔質ポリマーのフィルムまたは膜でもよく、例として多孔質ポリオレフィン、例えば多孔質ポリエチレン、多孔質ポリプロピレンや、多孔質フルオロポリマーや、または連続気泡、多孔質のポリウレタンといったといったものでもよい。更に、片方の面から他方の面につながる連続気孔を持つ薄い多孔質セラミックプレートのような、連続気泡、多孔質の無機材を使用できる。
多孔質フルオロポリマーには、多孔質ポリテトラフルオロエチレン（ＰＴＦＥ）、多孔質の延伸膨張ポリテトラフルオロエチレン（ｅＰＴＦＥ）、ポリテトラフルオロエチレンおよびポリエステルもしくはポリスチレンの多孔質共重合体、テトラフルオロエチレンおよびフッ素化エチレン−プロピレン（ＦＥＰ）もしくはＣ₁〜Ｃ₄のアルコキシ基を持つペルフルオロアルコキシ−テトラフルオロエチレン（ＰＦＡ）の共重合体が含まれるが、多孔質フルオロポリマーはこれらに限定されない。好ましい多孔質材料には、延伸膨張ポリプロピレン、多孔質ポリエチレンおよび多孔質ポリテトラフルオロエチレンが含まれる。最も好ましくは、この材料は、ノードがフィブリルで相互接続されたミクロ構造を持ち、気孔率が２０〜９０％の、延伸膨張ポリテトラフルオロエチレンであり、例えば参照してここに組み入れられる米国特許第３，９５３，５６６号明細書の教示に従って調整される材料のようなものである。
一般にＺ軸部材は約５〜５００μｍ、好ましくは約５〜１２５μｍの厚さを持つ。しかしながら、紫外光が試料を貫く限り厚さは決定的な要素にはならない。
図１２〜１４を参照して、Ｚ軸部材を形成するための材料１７０は、フィブリル２００で相互接続するノード１９０の間の空間として定義される気孔１８０を持つ延伸膨張ポリテトラフルオロエチレンである。フィブリル２００で相互接続するノード１９０の内部構造は、平板の部材の片方の側２３０から他方の側２４０へと、垂直に定められた断面２２０内のＺ軸を通る不規則な連続経路に、結果としてなる材料の密集体（ｍａｔｅｒｉａｌ ｄｅｎｓｉｔｙ）である。（図１４を参照）
米国特許第５，４９８，４６７号明細書により教示されるように、Ｚ軸方向に通る領域２２０を金属塩の付着を受けやすくして選択的に伝導性の部材２５０（図１４）を用意する。この金属塩は照射エネルギーを受けると非伝導性の金属核に変えられ、ついでこれが、無電解金属メッキ浴からの伝導性金属の析出を触媒するために働く。多孔質部材１７０の気孔１８０は初め、アルコール、または有機の水性界面活性剤のような湿潤剤で湿らせる。メタノール、プロパノール、テトラフルオロエチレン／ビニルアルコール共重合体またはその種の他のものが使用できる。湿潤剤は部材の材料が、ニッケルまたは銅のような伝導性金属を受け入れるように働く。特に好ましいのは銅である。
感照射性金属塩組成物は、感光性還元剤、金属塩、ハロゲン化イオン源、および第二還元剤（ｓｅｏｃｎｄ ｒｅｄｕｃｉｎｇ ａｇｅｎｔ）の溶液を含む液体の感照射性組成物である。感照射性溶液は、水、金属塩、感光性還元剤、第二還元剤を含有し、および任意に（表面をぬらしにくい場合に）界面活性剤を含有する。金属塩には、酢酸銅、蟻酸銅、臭化銅、硫酸銅、塩化銅、塩化ニッケル、硫化ニッケル、臭化ニッケル、第一鉄含有化合物、例えば硫酸第一鉄、塩化第一鉄等、並びにパラジウム、白金、銀、金、およびロジウムのような貴金属が含まれるが、含有塩はこれらに限定されない。
ふさわしい感光性還元剤は、芳香族ジアゾ化合物、鉄塩、例えばシュウ酸第二鉄または第一鉄、硫酸第二鉄アンモニウム等、ニクロム酸塩、例えばニクロム酸塩アンモニウム等、ジスルホン酸アントラキノンまたはその塩、グリシン（特に湿った表面条件で活性）、Ｌ−アスコルビン酸、アジ化物、および他の同様なものであり、また金属促進剤、例を挙げれば、例えば塩化第一スズといったスズ化合物、または銀、パラジウム、金、水銀、コバルト、ニッケル、亜鉛、鉄などの化合物であって、後者の群は１リットルに対して１ｍｇ〜２ｇの量で任意に加えられる。
第二還元剤には、グリセロール、エチレングリコール、ペンタエリトリトール、メソエリトリトール、１，３−プロパンジオール、ソルビトール、マンニトール、プロピレングリコール、１，２−ブタンジオール、ピコナール、スクロース、デキストリン、といったような多価アルコールや、トリエタノールアミン、プロピレンオキシド、ポリエチレングリコール、ラクトース、スターチ、エチレンオキシドおよびゼラチンなどの化合物が含まれるが、第二還元剤はこれらに限定されない。やはり第二還元剤として使用できる化合物は、ホルムアルデヒド、ベンズアルデヒド、アセトアルデヒド、ｎ−ブチルアルデヒドなどのアルデヒドや、ナイロン、アルブミンおよびゼラチンなどのトリアミドが含まれ、４−ジメチルアミノトリフェニルメタン、４’、４’、４’’−トリ−ジ−メチルアミノ−トリフェニルメタンなどのトリフェニルメタン染料のロイコ塩基や、３，６−ビスジメチルアミノキサンテンおよび３，６−ビスジメチルアミノ−９−（２−カルボキシエチル）キサンテンなどのキサンテン染料のロイコ塩基や、エチレングリコールジエチルエーテル、ジエチレングリコール、ジエチルエーテルテトラエチレングリコールジメチルエーテルなどのポリエーテルや、他の同様なものが含まれるが、第二還元剤として使用できる化合物はこれらに限定されない。
ソルビトールによって例証される保湿剤でもある第二還元剤は、明らかに現像する前に「乾燥」した被膜への調湿効果の理由から、１つの態様では処理溶液の保湿剤用の成分として利用される。それは、被膜中の全ての未転化の感照射性組成を塩基で洗い流す現像工程の間、部材の内部材料の金属被膜の密度を維持するのを実質的に補助する。
ふさわしい界面活性剤には、Ｒｏｈｍ＆Ｈａｓｓ Ｃｏ．製のＴＲＩＴＯＮ Ｘ−１００のようなポリエテノキシ非イオンエーテル、およびＯｌｉｎ Ｍａｔｈｉｅｓｏｎ Ｃｏｍｐａｎｙ製の界面活性剤６Ｇおよび１０Ｇのようなノニルフェノールとグリシドールの反応に基づく非イオン界面活性剤がある。
この処理溶液、すなわち、感照射性組成物は、通常水溶液のｐＨを２〜４（好ましくは２．５〜３．８）の間に調整するために酸性塩の形の酸性化剤、および少量のハロゲン化物イオン（ヨウ素、臭素または塩素イオン）含み、それにより添加剤の組み合わせが、処理した平板の材料を照射エネルギーにさらすことにより後で作られる被膜の密度を実質的に高めることに驚異的な効果を与える。酸性度を調整することはその目的のためだけの薬剤を導入することを必ずしも必要とするとは限らない。なぜなら、調整は、酸性の性質の感光性還元剤（例えば、アスコルビン酸、グリセリンなど）またはハロゲン化物イオンを導入するための他の何らかの添加剤（例えば塩酸）で例証されるような、他の機能も持った酸性の物質により、完全にまたは部分的になされることがあるからである。同様に、一部のまたは全てのハロゲン化物イオンを、還元できる金属塩（例えば、第二銅）の成分として導入してもよい。
感受性にした溶液のｐＨを制御または調節するために使用できる多くのふさわしい酸性物質のうちには、フルオロホウ酸、クエン酸、乳酸、リン酸、硫酸、酢酸、蟻酸、ホウ酸、塩酸、硝酸および他の同様なものがある。幅広い種類の臭化物、塩化物及びヨウ化物塩およびその他のハロゲンを発生させる水溶性の化合物を利用して、処理溶液の要求されるハロゲン化物イオン含有量の一部または全てを供給してもよい。これらにはとりわけ、一般に金属とこれらのハロゲンとの塩で、臭化第二銅、塩化ニッケル、塩化コバルト、塩第二銅、ヨウ化ナトリウム、ヨウ化カリウム、塩化リチウム、ヨウ化マグネシウム、臭化マグネシウム、臭化ナトリウム、臭化カリウム、および他の同様なもので例証されるようなものを含めることができる。臭化物塩は、少なくとも特定の場合に基板上での感度を対応する塩化物よりも高くする（すなわち、より濃く密な析出物）生じるので、より好ましい。
ハロゲン化物イオンは溶質のうちより少ない割合のみを構成し、典型的には、溶解固形分の全重量基準で、約０．０４５〜１．６％、好ましくは約０．１３〜０．４５％の範囲でよい。ハロゲンの量は他の表し方をすれば、感受性溶液１リットル当たりのハロゲンが約０．９〜２５ミリ当量、好ましくは２．５〜９ミリ当量、例えば臭化第二銅では０．３〜１．０ｇ／リットルにしてよい。ハロゲン化物イオンの割合を増加させることは、そのような増加が処理の感受効果を最適な量で得られるもの未満に徐々に減少させるので、通常好ましくない。また、当量として表されるこれらのハロゲン化物イオンの割合は、処理溶液の第二銅または他の還元性非貴金属カチオンのそれよりも少ない。例えば、そのような金属イオンの当量のハロゲン化物イオン当量に対する比は、通常少なくとも２：１，好ましくは４：１〜１００：１の範囲である。
感照射性組成物を材料１７０に適用して気孔を画成している材料を完全にぬらし、それにより多孔質部材を、組成物が材料１７０の気孔１８０を透過しまたは貫通して、多孔質の平面材料の片方の面から他方の面までの気孔を画成する材料に沿って気孔の内部に被膜を形成するのに十分な時間、感照射性組成物にさらす。この後、被覆した多孔質部材を自然乾燥または５０℃以下での炉での加熱で乾燥する。この段階で、処理用組成物の感光性を保持するために、材料は黄色光条件下で処理すべきである。また、処理に不利な影響をおよぼしかねない部材材料による水の吸収の可能性があるので、部材を７０°Ｆより低い温度、６０％以下の相対湿度に保つべきである。
被覆した多孔質部材の片方の表面の選ばれた場所を不透明の覆い２６０（図１２）でマスクし、カバーした領域にその後の照射線が当たらないようにする。マスク２６０は結果として、非伝導性領域の一つおきのバンドにより切離されＺ軸方向に通り抜ける任意の、所望の、形、大きさ、配列の点状の伝導性領域、または伝導性領域の交互のバンドまたはストリップをもたらすことができる。点は通常円形であるが、正方形、長方形などの他の幾何学的な形状であることができよう。点の大きさは小さくて０．０００１インチ、大きくて０．０２５インチほどでよく、好ましくは０．００１インチ、０．００２インチ、０．００３インチ、０．００４インチ、０．００５インチ、０．００８インチ、０．００９インチまたはこれらの端数（ｆｒａｃｔｉｏｎ）でよく、ここで、隣接する点の中心間の距離として定義されるピッチは、好ましくは点の寸法の少なくとも２倍で、例えば１ｍｉｌの点では、２ｍｉｌのピッチである。
マスクされた部材１７０を、光、電子ビーム、ｘ線、および同様なもの等の照射線、好ましくは紫外線に、金属塩の金属カチオンを部材の肉厚を通して還元して金属核にするのに十分な時間かつ十分なパワーで暴露する。その後で部材１７０のマスクを取り除き、酸性またはアルカリ性の洗浄溶液で洗浄して、不透明なカバーで保護されていた感照射性組成物を洗い流す。酸性またはアルカリ性洗浄（または定着）溶液を数分より長く、例えば５分以下の間、接触したままにしなければ、溶液は照射線が金属カチオンを金属核に還元した領域に影響を及ばさない。
具体的には、図１２で示すように、処理される部材１７０は金属マスク、ジアゾまたはハロゲン化銀フィルムで選択的にマスクされる。マスクした部材１７０をその後、５００ｎｍより短い波長の視準されていないまたは視準された紫外光源で、写真描写する。安定なホトイメージを作るために、触媒、非伝導性の金属核、それ自身は最小で２００ｍＪの照射エネルギーを必要とする。
この紫外光エネルギーは、多孔質部材の肉厚を貫くのに十分強力である。この様に、続くメッキ工程では、伝導性の金属が連続的にＺ軸を通してメッキして、Ｚ軸に電気的な連続性を与える。必要なら、紫外光エネルギーを平面部材の両面に当てることができる。
５分間のならし（ｎｏｒｍａｌｉｚａｔｉｏｎ）期間後、触媒された材料を３０〜９０秒の短時間、硫酸溶液で、例えば８重量％の硫酸と９２重量％の脱イオン水からなる溶液で、または水酸化ナトリウムで１０より高いｐＨに調節した、４０ｇ／リットルのエチレンジアミン四酢酸、１００ｍｌ／リットルのホルムアルデヒドからなるアルカリ溶液で洗浄する。この洗浄工程の目的は、光還元像を保持しながら材料から露光されなかった触媒を除去することである。
選択的な像を含む洗浄された材料を、次に反応性金属カチオン置換溶液で安定化する。

像はより安定なカチオン、例えばパラジウムによる銅の置換反応を受ける。この様に薄い層の量では銅がより速く酸化する傾向があり、パラジウムが無電解液中での還元反応をより速く開始させる能力もあることから、より安定な系が所望される。部材をこの溶液中に少なくとも３０秒とどめ、次に脱イオン水中で約１分間洗浄する。
触媒反応を受けた部材を選択的に、１またはそれ以上の伝導性金属で約５０〜６０マイクロインチの析出厚さに無電解メッキする。この様な金属には銅、ニッケル、金、銀、白金、コバルト、パラジウム、ロジウム、アルミニウムおよびクロムが含まれる。電解浴に入れられている間、部材を揺動運動させて撹拌し、基板の最も奥の領域への金属の拡散を促進する。メッキは、最初に脱イオン水中ですすぎを行い、その後で材料中でパラジウムの上に、かつ基板の肉厚を通して銅を析出させるのに十分な時間、撹拌している無電解銅浴に浸漬して行われる。
この様に、Ｚ軸方向に材料を貫く選択された領域内で、材料のノード１９０およびフィブリル２００を、Ｚ軸部２９０並びに上部および下部の接触パッド２７０および２８０を持つ伝導性金属層２３０で、少なくとも部分的に覆う。伝導性の金属は、上部および下部のパッド２７０および２８０の間の選択された領域を通る連続の伝導経路２９０を形成する。
以下の例では、使用する触媒処理溶液を１リットルの脱イオン水に下記のものを加えて調整した。

使用した定着溶液は８重量％の硫酸、９２重量％の蒸留水であった。安定剤も使用し、以下の成分を含む。

例９
米国特許第５，４９８，４６７号明細書により教示されるように、Ｗ．Ｌ．Ｇｏｒｅ ＆ Ａｓｓｏｃｉａｔｅｓから得られる伸張した多孔質ポリテトラフルオロエチレン膜を、室温で約３０分間メタノール７５％、エタノール２５％およびテトラフルオロエチレンとビニルアルコールの共重合体１重量％の溶液に浸漬することにより、湿潤剤で処理した。
湿潤した膜をその後、触媒処理溶液に６０秒間浸漬し、その後５０℃の炉の中で３分間乾燥した。その後、膜の片方の表面を直径６ｍｉｌおよび６ｍｉｌピッチ（中心から中心まで）の点を持つジアゾフィルムでマスクした。
その後、膜を１６００ｍＪの視準された紫外光源に約２分間暴露した。５分間のならし期間の後、紫外光で処理した膜を、３０秒間定着溶液中で洗浄し、暴露されなかった触媒処理溶液を除去した。次に、選択的に描画した膜を１分間安定化溶液に浸漬して安定化し、その後蒸留水で１分間洗浄した。
安定化した膜をその後、脱イオン水１リットル当たりを基準として、

の銅メッキ浴組成物（Ｓｈｉｐｌｅｙｓ ３）に浸漬した。
膜を７．５分間撹拌棒を使用してこの浴の中で撹拌し、銅がＺ軸を通る触媒された部分の膜の気孔の全体にわたって拡散するのを促進した。
伝導性のＺ軸材の経路に、２つの他の伝導性材料の間のコネクター界面として機能するよう接着剤を充填することができる。ふさわしい接着剤には、エポキシ樹脂、アクリル樹脂、ウレタン樹脂、シリコーン樹脂、ポリイミド樹脂、シアネートエステル樹脂、その他同様なものが含まれる。通常、接着剤は膜を接着剤の溶液中に浸漬して気孔へ吸収させている。エポキシ樹脂の場合には、ふさわしい溶媒はメチルエチルケトンである。接着剤を伝導性のＺ軸材に吸収または含浸させたなら、結合能力を与えるために、それを１６０℃で焼成してＢ段階の伝導性Ｚ軸部材を作る。
例１０
６ｍｉｌ（１５０μｍ）の厚さの、伸張した、多孔質ポリテトラフルオロエチレン膜を、２プロパノールに１分間浸漬して湿潤させた。その後、例９のようにそれを触媒処理溶液に１分間浸漬して、乾燥し、マスクし、紫外光に暴露した。それをその後、例９の様に定着および安定化溶液で処理し、次に例９のように銅でメッキした。
例１１
２ｍｉｌの厚さの伸張した多孔質ポリテトラフルオロエチレン膜を、マスクするストリップを１０ｍｉｌピッチの３ｍｉｌのパッドにすることを除いて、例９と同様に調整した。
例１２〜１４
Ｚ軸に沿って延びる不規則な形の伝導性経路を持つ伝導性Ｚ軸材料を作るため、多孔質ポリエチレン、多孔質ポリプロピレンおよび連続気孔の多孔質ポリウレタンから作られた膜について、例９の手順を踏襲した。
例１５
図１５（倍率１０００倍）で示すノード−フィブリル構造を持つ伸張した多孔質ポリテトラフルオロエチレン膜から作った膜は、２５℃で密度０．２２ｇ／ｃｍ³および空気容積７０％の厚さ７６μｍのもので、Ｗ．Ｌ．Ｇｏｒｅ ＆ Ａｓｓｏｃｉａｔｅｓより入手可能であって、伝導性Ｚ軸材を作るためにマスクするストリップを５ｍｉｌピッチの２ｍｉｌのパッドにすることを除いて例９の様に調整された。
例１６
Ｗ．Ｌ．Ｇｏｒｅ ＆ Ａｓｓｏｃｉａｔｅｓにより入手可能な、２５℃で密度０．４ｇ／ｃｍ³および空気容積２０％の厚さ４０μｍの、図１６（倍率１５００倍）で示すノード−フィブリル構造を持つ伸張した多孔質ポリテトラフルオロエチレン膜は、伝導性Ｚ軸膜を作るために、マスクするストリップを１５ｍｉｌピッチの８ｍｉｌパッドにすることを除いて例９の様に調整された。
例１７
２５℃で密度０．３５ｇ／ｃｍ³および空気容積３０％の厚さ１００μｍの、図１７（倍率１０００倍）のノードフィブリル構造を持つ伸張した多孔質ポリテトラフルオロエチレン膜は、Ｗ．Ｌ．Ｇｏｒｅ ＆ Ａｓｓｏｃｉａｔｅｓにより入手可能で、伝導性のＺ軸材を作るために、マスクするストリップを１５ｍｉｌピッチの８ｍｉｌパッドにすることを除いて、例９の様に調整された。
例１８
２５℃で密度０．２０ｇ／ｃｍ³および空気容積７０％の厚さ１５０μｍの、図１８（倍率１０００倍）のノードフィブリル構造を持つ伸張した多孔質ポリテトラフルオロエチレン膜は、Ｗ．Ｌ．Ｇｏｒｅ ＆ Ａｓｓｏｃｉａｔｅｓにより入手可能で、伝導性のＺ軸材を作るために、マスクするストリップを１５ｍｉｌピッチの８ｍｉｌパッドにすることを除いて、例１の様に調整した。
上で説明した伝導性の経路を含む例１０〜１８の伝導性Ｚ軸材料に、２つの他の伝導性材料間のコネクター界面となるように接着剤を充填することができる。ふさわしい接着剤にはエポキシ樹脂、アクリル樹脂、ウレタン樹脂、シリコーン樹脂、ポリイミド樹脂、シアネートエステル樹脂、または同様なものが含まれる。通常、接着剤は接着剤溶液中に膜を浸すことで気孔へ吸収させている。エポキシ樹脂について言えば、ふさわしい溶媒はメチルエチルケトンである。例９のように、結合能力を与えるためのＺ軸材料にエポキシ接着剤を含浸させて、１６０℃で焼成してＢ段階のＺ軸複合材料を供給する。
ここでは本発明の特定の態様について図示および説明したが、本発明は図示および説明したようなものに限定されない。請求の範囲の記載の請求項の範囲内で変更や改変を本発明の一部として取り入れ、具体化してもよいことは明らかである。Field of Invention
The present invention is directed to a dimensionally stable core for high density chip packages that reduces the variability of material movement during processing and chip package manufacture. The stable core that provides stability to the high-density chip package is preferably made of metal and includes a dielectric layer and a metal layer formed simultaneously. Subsequently, vias are formed to enable the manufacture of high density chip packages.
Background of the Invention
High density chip package substrates require many vias that connect the input / output (I / O) of the integrated circuit chip to a routing layer or power / ground layer embedded in the chip package substrate. Traditional mechanical drilling processes used in the past for via formation have been replaced by laser via drilling processes. The laser via drilling process offers the potential to form vias much faster at a lower cost than conventional mechanical drilling processes. In order to allow this fast rate of via formation, it is advantageous to use a thinner dielectric than has been used previously. Thin dielectrics allow the formation of microvias, ie vias <100 μm in diameter.
After forming these microvias, the aspect ratio (via height divided by its diameter) must be kept at a predetermined value in order to effectively metallize them. For example, in the case of closed vias, the aspect ratio should be approximately 1 to ensure that conventional metallization methods are adequate to deposit a sufficient amount of metal on the microvia wall.
As should be appreciated, it is disadvantageous to form micro one-closed vias in dielectric materials including glass fabric reinforcements. More specifically, using a glass fabric reinforced dielectric, the laser's ability to remove the glass fabric is necessary to remove the embedded copper pad on the layer to which the closure via is to be connected. It becomes as big as the ability of the laser. Thus, the laser penetrates the copper pad and reaches the layer below, thus expanding the via. Unintentionally expanded vias can short circuit to subsequent layers after metallization and scrap the entire circuit.
In order to eliminate the disadvantages associated with forming micro one-closed vias in dielectric materials including glass fabrics, dielectric material manufacturers have introduced thin, unreinforced prepregs instead of conventional glass reinforced prepregs. . However, the use of such thin, unreinforced prepregs has created other difficulties in the manufacture of high density printed circuit boards and chip package substrates. More specifically, removing the glass fabric from the dielectric material reduced the dimensional stability of the chip package substrate. When thermosetting resins are used as the constituent material of the dielectric material, these resins generally induce shrinkage of the multilayer circuit board or the chip package substrate due to the lamination and curing of the resins. This shrinkage is about several thousand ppm at most, and since the shrinkage can be compensated for during the production of the substrate, it is not a manufacturing problem in itself. However, even one invariant circuit design is variable in the degree of shrinkage. This variability in the degree of shrinkage can lead to lower yields of high density printed circuit boards and chip package substrates. For example, if one of the inner layers of a multilayer circuit is a power or ground layer consisting of a copper plane, and the vias must pass through the clearance of this layer, the variability in shrinkage caused during lamination is from the panel. This may change the position of the ground clearance to the panel. When drilling vias in these circuits, these vias may lose ground or power layer clearance at some rate and collide with copper. After metallization, these vias will short to the ground or power layer, resulting in scrap circuitry.
There is a need for a dimensionally stable core to facilitate the manufacture of high density chip packages with dielectric materials that are not supported by a glass fabric.
Summary of the Invention
The present invention includes a method of manufacturing a high density chip package substrate by simultaneously forming circuit layers on both sides of a metal core, preferably made of copper. The copper core improves the dimensional stability of the unreinforced thin dielectric material used in the substrate. The metal core reduces the movement of the material during stacking, thereby facilitating the processing and manufacturing of high density chip packages. The metal core also provides an embedded capacitance that helps reduce the simultaneous switching noise of the chip.
The dimensionally stable core of the present invention comprises a metal core with at least one clearance etched through it, a dielectric layer on both sides of the metal core and filling the clearance, and a metal cap layer on both sides of the dielectric layer And vias drilled through the metal cap layer and through the metal core. When electrically insulating a metal core, vias are made to extend through the clearance of the metal cap layer, dielectric layer and metal core, thus isolating the metal core. A metal layer is attached to the wall surface of such a via. The dielectric layer and the metal cap layer can be simultaneously disposed on the upper surface and the lower surface of the metal core. Additional dielectric layers and metal cap layers can also be laminated simultaneously on the top and bottom surfaces of the cap in a similar manner. Form additional vias if necessary. By repeatedly adding dielectric and metal cap layers and vias, a dimensionally stable high density chip package is provided.
The metal core and metal layer may be copper or any other suitable metal such that the coefficient of thermal expansion (CTE) of the chip package matches the CTE of the printed circuit board. The dielectric may be any unsupported prepreg. In addition, the vias can be occluded vias or through vias. Vias are drilled using a laser that removes material when drilled.
A method of manufacturing a dimensionally stable core according to the present invention supplies a metal core such as copper, disposes a dielectric layer on the upper and lower surfaces of the metal core, and attaches a metal cap layer such as copper to the dielectric layer. And laminating the dielectric layer and the metal cap layer together on the metal core. The method further includes drilling the vias through the metal cap layer, dielectric layer and metal core, or metal core clearance, and then repeating the above steps to produce a multi-layer high density chip package. Including.
[Brief description of the drawings]
The foregoing summary, as well as the following detailed description of preferred embodiments of the invention, will be better understood when read in conjunction with the appended drawings. For the purpose of illustrating the invention, there are shown in the drawings preferred embodiments which are presently preferred. It should be understood, however, that the invention is not limited to the precise arrangements and instrumentalities shown. In the drawing
FIG. 1 is a side view of a metal core before processing.
FIG. 2 is a side view of a metal core having a patterned photoresist before etching.
FIG. 3 is a side view of the metal core of FIG. 1 with clearance etched through.
FIG. 4 is a side view of the substrate after laminating a dielectric layer and a metal “cap” layer on a metal core.
FIG. 5 is a side view of a substrate according to the present invention after vias have been drilled with a laser, metallized, and circuit traces have been formed.
FIG. 6 is a side view of the substrate of FIG. 4 after adding additional dielectric layers and metal “cap” layers.
FIG. 7 is a side view of a completed chip package according to the present invention after one-sided vias have been drilled.
FIG. 8 is a side view of a substrate according to a preferred embodiment of the present invention.
FIG. 9 is a scanning electron micrograph of an expanded polytetrafluoroethylene (ePTFE) material that can be used as a dielectric layer in the present invention.
FIG. 10 is a scanning electron micrograph of an expanded PTFE material forming a matrix containing an adhesive filler in order to form one dielectric material in the present invention.
FIG. 11 shows two substrates laminated together using a conductive Z-axis material.
FIGS. 12-14 schematically illustrate the structure of a conductive Z-axis member including irregularly shaped conductive paths.
FIGS. 15-18 are scanning electron micrographs showing some ePTFE matrix node fibril structures used for making conductive Z-axis materials.
FIG. 19 shows two substrates including one closed vias laminated with conductive Z-axis material.
Detailed Description of the Invention
The present invention is directed to a dimensionally stable core for high density chip packages. This stable core is a metal core with clearance, preferably copper. Dielectric layers are disposed on the upper and lower surfaces of the metal core. The metal cap layer is disposed on each of the upper surface and the lower surface of the dielectric layer. The dielectric layer and the metal cap layer can be simultaneously disposed on both sides of the package. One closed via or through via is then drilled through and through the metal cap layer to the dielectric layer and metal core clearance. The laser is used to drill the vias by removing the material where the vias are to be made. If the metal core is a power or ground plane, the via is made directly through the metal core. If the metal core is an insulated metal core, then the vias extend through the clearance of the metal core, i.e. the metal core does not have vias in direct contact with them.
During the chip package stacking process, the stable core reduces the movement of the substrate material and achieves uniform shrinkage from substrate to substrate. This allows each substrate to perform the same function. Furthermore, a plurality of dimensionally stable core chip packages can be joined together to obtain a high density chip package.
As can be seen in FIG. 1, the metal core 10 is a metal sheet, such as a copper sheet, with rough surfaces 12 and 14 on opposite sides of the metal. This metal sheet has been processed in a process known as the “double treat” method. This rough surface allows for better adhesion of the dielectric layer. Other metals can be used in place of copper as long as the coefficient of thermal expansion (CTE) of the metal core closely matches the CTE of the printed circuit board to which the chip package will eventually be attached. For example, CIC, a copper-INVAR 36-copper laminate, can be used in place of an integral copper core. The nickel alloy used can contain 30% to 42% by weight of nickel. The metal core 10 is thick enough to provide dimensional stability to the finished substrate, for example 0.0005 inch to 0.0028 inch (12.7 μm to 71.1 μm), preferably 0.0014 inch (35 .6 μm). A thickness in this range gives the chip package a lower flexural modulus.
As shown in FIG. 2, photoresist 22 is applied to both sides of the metal core 10 using conventional methods to draw a pattern of clearances 20 in which vias are to be formed. The drawn location is chemically etched by conventional methods to remove the copper and create the clearance 20 (FIG. 3). This clearance 20 is larger in diameter than a via that will be produced later. Via dimensions depend on the type of dielectric placed on the metal core. The photoresist is removed in the usual way.
As shown in FIG. 4, the metal core 10 including at least one clearance 20 includes dielectric layers 30 and 32. Preferably, the dielectric layers 30 and 32 are formed simultaneously on the surfaces 12 and 14 of the metal core 10. The CTE of dielectric layers 30 and 32 can be as high as 40 ppm / ° C. Dielectric layers 30 and 32 must be thick enough to satisfy clearance 20. The thickness of the dielectric layers 30 and 32 is at least 5 μm or more, and typically 25 μm to 50 μm. The thickness is chosen so as not to lose the dimensional stability of the substrate. The dielectric thickness should not be significantly greater than the thickness of the metal core 10 or the dimensional stability of the substrate will be reduced.
Metal core 10 includes metal cap layers 40 and 42 on surfaces 34 and 36 of dielectric layers 30 and 32, respectively. Metal cap layers 40 and 42 are simultaneously disposed on the surfaces 34 and 36 of the dielectric layer. This metal cap layer is made of copper. The metal cap layers 40 and 42 can be made of any suitable conductive metal. Metal core 10, dielectric layers 30 and 32, and metal cap layers 40 and 42 are laminated together using any suitable lamination method.
In one embodiment of the invention, layers 30 and 32 are made of W.W. L. Made from SPEEDBOARD C ™ from Gore and Associates. Dielectric layers 30 and 32 were laminated to core 10 at a pressure of 300-350 psi and a temperature increase rate of 5-7 ° C. per minute until a temperature of 177 ° C. was reached. The temperature was kept at 177 ° C. for 30 minutes. The temperature was then raised to 220-225 ° C. and maintained for 60 minutes. The laminate was slowly cooled over several hours while pressing the substrate.
As shown in FIG. 5, the laminated dielectric layers 30 and 32 and the metal cap layers 40 and 42 include at least one via 50 drilled through the metal cap layers 40 and 42 and the dielectric layers 30 and 32. To be processed.
FIG. 5 shows the via 50 as a through via that can be drilled using any known method, such as a laser. In the case of through vias 50, the laser must be able to penetrate the dielectric layers 30 and 32 and the metal core 10 when not passing through the clearance 20. The aspect ratio of the through via 50 can be 1: 1 to 20: 1 or higher. On the other hand, when making a closed via, the laser must penetrate the dielectric layer 30 or 32, but not the metal core 10. An atypical 1: 1 aspect ratio should be maintained. However, the aspect ratio is not limited to 1: 1.
According to the present invention, in order to perform uniform laser via formation with an appropriate aspect ratio, the number of pulse repetitions for changing the pulse width of the laser is set to a range of 1 kHz to 10 MHz. Residues remaining in the via are then removed again by adjusting the intensity of the pulse. Energy density 0.5 ~ 20J / cm²A laser is used that can change the energy density in situ and virtually instantaneously by changing the number of pulse repetitions from 1 kHz to 10 MHz.
In one embodiment of the present invention as in FIG. 7, the occluded via was made with a solid pulsed laser, such as a pulsed Nd: YAG laser. The basic output of the Nd: YAG laser is at a wavelength of 1064 nm. This wavelength is the infrared part of the electromagnetic spectrum. By introducing β-barium borate (BBO) crystal into the optical path, harmonic generation promotes the output of light at 355 nm (third harmonic) and 266 nm (fourth harmonic) entering the ultraviolet region. The wavelengths of 355 nm and 266 nm are particularly suitable for drilling vias in the laminated substrate of the present invention.
In general, the pulse length, energy density, and number of pulses can vary depending on the type of via being made and the type of material used in the laminated substrate. Using a 266 nm laser on a laminated substrate made of alternating layers of copper and dielectric layers composed of adhesive, filler and expanded polytetrafluoroethylene (ePTFE), one closed via When drilling holes, the energy density is 1.5 J / cm.²Energy per pulse is 10 J, and power density is 20 MW / cm²It is. At a wavelength of 355 nm, the energy density is 3.5 J / cm.²Energy per pulse is 30 J, and power density is 35 MW / cm²On the other hand, a closed via is made. In the process after the pulse of the closed via using a 355 nm laser, the energy density is 11 J / cm.²Energy per pulse is 100 J, and power density is 200 MW / cm²It is necessary to adjust the conditions so that
When using a 355 nm laser in adhesive-filler-ePTFE dielectric material while forming a closed via, the energy density is 7 J / cm.²Energy per pulse is 65J, and power density is 100MW / cm²It is. In the process after the pulse, these parameters are set so that the energy density is 11 J / cm.²Energy per pulse is 100 J, and power density is 200 MW / cm²It will be necessary to adjust so that
As shown in FIG. 5, the through via 50 is metalized by forming a metal layer 60 on the surface of the via 50. The metal layer 60 includes, for example, an electroless copper layer formed on the surface of the via 50 using an ordinary method, and an electrolytic copper layer formed on the electroless copper layer. Metal layer 60 extends over metal cap layers 40 and 42 and dielectric layers 30 and 32. Traces (not shown) or signal wirings are then etched into the metal cap layers 40, 42 and the metal layer 60 using techniques known in the art. That is, in the case of an inner layer trace, that is, a trace that is not on the last layer, the top metal cap layers 40 and 42 and the metal layer 60 are drawn using a photoresist. The photoresist is patterned and then developed so that the photoresist defines a signal trace in the area under the photoresist. The metal between the patterned photoresist is then etched to remove all remaining photoresist leaving traces around the vias. This is called a subtractive method.
For the trace of the last layer, that is, the trace on the last layer, the top metal cap layers 40 and 42 and the metal layer 60 are drawn with photoresist. The photoresist is then patterned and developed to remove the photoresist where signal traces are made. Where the photoresist is removed, an additional copper-like metal is plated. Thereafter, an etch-resistant metal such as nickel, gold, tin or solder is plated on the additional metal, the photoresist is removed, and the metal cap layers 40, 42 under the photoresist are etched. The areas to be left are plated with metal and the areas to be etched are exposed. This is called a semi-additive process.
Preferably, as much of the metal layer 60 as possible should remain on the dielectric layers 30 and 32 to further assist in making a dimensionally stable substrate. If the metal layer 60 is a ground or power plane, the operation of the method employed will leave a large amount of metal such as copper.
FIG. 6 illustrates another aspect of the present invention, where the metal cap layers 40, 42 and the metal layer 60 are not the last layer. In this embodiment, additional dielectric layers 70 and 72 are disposed on surfaces 62 and 64 of metal layer 60, respectively. Additional dielectric layers 70 and 72 fill via 50 and extend over metal layer 60, dielectric layers 30 and 32, and remaining metal cap layers 40 and 42. Additional dielectric layers 70 and 72 encapsulate the traces.
Additional metal cap layers 80 and 82, such as copper, are disposed on the surfaces 74 and 76 of the dielectric layers 70 and 72, respectively. As with the additional dielectric layers 70 and 72, the additional metal cap layers 80 and 82 can be placed on both sides of the package simultaneously. At that time, dielectric layers 70 and 72 and metal cap layers 80 and 82 are laminated together as described above.
FIG. 7 shows a mode in which a closed via is added to the substrate of FIG. On the other hand, closed vias 90 and 92 are made in metal cap layers 80 and 82 and dielectric layers 70 and 72, respectively, using a laser drilling technique as described above. The vias 90 and 92 are shown as closed vias, but may be through vias that extend completely through the substrate and connect any layer in the substrate to any other layer. . The vias 90 and 92 are then metallized by making a metal layer 100 using a suitable plating method. Occluded vias 90 and 92, on the other hand, have pads 100 that can be further processed to have metal traces (not shown) using techniques known in the art. To create 7, 9 or more layers, it is possible to continue adding dielectric layers and cap layers simultaneously or otherwise, as listed above.
It should be understood that the preferred chip package and the method of manufacturing the preferred chip package have been described. However, for example, the dielectric layer and the metal layer can be disposed one layer at a time on the substrate rather than simultaneously.
In another embodiment of the present invention, a substrate such as that shown in FIG. 5 or 7 or others including 3, 5, 7 or more layers as previously shown is very high. They can be joined together to make a dense chip package or a very dense printed circuit board. Each substrate layer is bonded together using a conductive Z-axis material.
For a chip package having a copper metal core with a dielectric layer and a metal cap layer made at the same time as listed above, the chip package is approximately 300-350 psi until a temperature of 177 ° C. is reached. A temperature rise rate of 5-7 ° C./min is applied. The temperature is maintained at 177 ° C. for approximately 30 minutes. The temperature is then raised to approximately 220 ° C. to 225 ° C. for approximately 60 minutes. The chip package is then slowly cooled under a pressure of about 300-350 psi maintained throughout the bonding process. Allow several hours for cooling.
FIG. 8 shows a chip package according to the present invention having an electrically insulated copper core 110 and prepreg sheet 120. The prepreg sheet 120 is named WPE under the name SPEEDBOARD ™. L. Manufactured by Gore & Associates. Such a prepreg sheet has a dielectric constant of 2.6 to 2.7, a glass transition temperature (Tg) of 220 ° C., a fluidity (fIow) of less than 4%, and a curing temperature of 177 ° C. to 225. ° C. The prepreg sheet 120 has a thickness in the range of about 25 μm to about 175 μm.
The substrate of FIG. 8 was evaluated and compared with a system using a conventional polyimide resin core. The goal of optical alignment using the substrate of FIG. 8 is measured after the final stacking process and compared to a chip package with a substrate containing a conventional polyimide core, the dimensionality of each chip package. Variability was sought. It was found that the standard deviation of shrinkage after lamination was about 200 ppm in the conventional polyimide resin core structure. It has been found that for chip packages made in accordance with the teachings of the present invention, the standard deviation of shrinkage after stacking is only about 25 ppm.
For high density chip packages, such a substantial reduction in post-stack shrinkage standard deviation can significantly reduce scrap due to shorted or open vias. Since the effect of the standard deviation of shrinkage after large stacking is greater at the edge of the printed circuit board, it also allows the assembly of the substrate with larger sized panels.
In the above example, a specific dielectric material was used to make the package. However, in the present invention, any suitable dielectric material, such as polyimide resins, and polyimide laminates, epoxy resins, epoxy resins combined with other resin materials, organic materials, alone or in combination with fillers Although any of the above can be used, suitable dielectric materials are not limited to these. Preferred dielectrics include fluoropolymer matrix, expanded polytetrafluoroethylene (ePTFE), or copolymers or blends, where the fluoropolymer can be polytetrafluoroethylene (PTFE). Suitable fluoropolymers include, but are not limited to, polytetrafluoroethylene with or without an adhesive filler mixture or expanded polytetrafluoroethylene.
Preferred materials include W.P., such as SPEEDBOARD ™ C, a prepreg of a nonwoven material comprising a cyanate ester resin in a polytetrafluoroethylene matrix. L. SPEEDBOARD ™ bond ply available from Gore & Associates is included. SPEEDBOARD ™ C has a dielectric constant (Dk) of 2.6 to 2.7, a loss of tangent of 0.004, a dielectric strength of 1000 V / mil, a glass transition temperature (T_g) Is 220 ° C. and the resin content is 66 to 68% by weight, and is available in various thicknesses.
Another suitable dielectric is the expanded PTFE matrix shown in FIG. 9, which comprises a mixture of at least two of a cyanate ester compound, an epoxy compound, a bistriazine compound and a poly (bis-maleimide) resin. For example, the varnish solution is 5.95 pounds of M-30 (Ciba Geigy), 4.07 pounds of RSL 1462 (Shell Resins, Inc), tetrabromobisphenol A carbonate terminated with 2,4,6-tribromophenyl. 4.57 pounds of oligomer (BC-58) (Great Lakes Inc.), 136 g of bisphenol A (Aldrich Company), 23.4 g of Irganox 1010, 18.1 g of Mn HEX-CEM 10% mineral spirit solution, and MEK 8. Made by mixing 40 kg. The varnish solution was further diluted to two separate baths, 20% (w / w) and 53.8% (w / w). The two varnish solutions were poured into separate impregnation baths and the e-PTFE web was passed through each impregnation bath immediately after passing one through the other. The varnish was constantly stirred to compensate for uniformity. The impregnated web was then immediately passed through a heated oven to remove all or almost all of the solvent and the adhesive was partially cured and collected on a roll. The ePTFE web can be of any desired thickness, for example 25 μm, 40 μm. The material having a thickness of 25 μm has a mass of about 0.9 g and a weight per unit area of about 11.2 to 13.8 g / m.²It is.
Another aspect of the invention is defined by dielectric layers 30, 32, 70 and 72 made from a porous matrix system that absorbs or impregnates the adhesive-filler mixture. FIG. 10 shows an ePTFE matrix node-fibril substructure illustrating one embodiment of such a porous matrix system.
In general, a porous matrix in the present invention is a nonwoven substrate that absorbs a large amount of filler and thermoplastic or thermosetting adhesive as a result of the initial porosity of the substrate. Such porous matrix is then heated to partially cure the adhesive and form a B-stage composite. The substrate comprises a fluoropolymer such as the porous expanded polytetrafluoroethylene material of US Pat. Nos. 3,953,566 and 4,482,516, incorporated herein by reference. Preferably, the mean flow pore size (MFPS) should be about 2-5 times greater than the largest particles, with MFPS being more than approximately 2.4 times larger than the largest particles. However, it is still within the scope of the present invention that a suitable composite can be tuned by selecting a ratio such that the average flow pore size to the average particle size is greater than 1.4. Acceptable composites can also be adjusted when the ratio of minimum pore size to average particle size is at least greater than 0.8, or when the ratio of minimum pore size to maximum particle size is at least greater than 0.4. Such ratios can be examined with the Microtrak ™ model FRA Particle Analyzer.
Alternatively, another method for measuring relative pore size and particle size may be calculated assuming that the minimum pore size is greater than or equal to about 1.4 times the maximum particle size.
In addition to the expanded fluoropolymer substrate described above, ultra-high molecular weight (UHMW) polyethylene, a porous expanded polyolefin such as expanded polypropylene, or polytetrafluoroethylene that is prepared by paste extrusion and incorporates a sacrificial filler, Porous inorganic or organic foams and microporous cellulose acetate can also be used.
The initial porosity of the porous substrate is at least 30%, preferably at least 50%, most preferably at least 70%, and this substrate is intended to prevent brittleness of the overall composite material and settling of particles. While providing a flexible reinforcement, it facilitates the impregnation of the pores with a thermosetting or thermoplastic adhesive resin and particulate filler paste.
The filler includes a collection of particles representing the maximum particle size, the minimum particle size, and the average particle size according to a histogram when analyzed with a Microtrak ™ model FRA Partial Analyzer device.
Suitable fillers for incorporation into the adhesive include BaTiO_Three, SiO₂, Al₂O_Three, ZnO, ZrO₂TiO₂Suitable fillers include, but are not limited to, precipitated ceramics and sol-gel ceramics such as silica, titania and alumina, non-conductive carbon (carbon black) and mixtures thereof. Particularly preferred fillers are SiO₂, ZrO₂TiO₂Are used alone or in combination with non-conductive carbon. Most preferred fillers include fillers made by the steam metal combustion method taught in US Pat. No. 4,705,762, such as a uniform surface curvature that is essentially solid, i.e., not a hollow sphere. And silicon, titanium, and aluminum to make silica, titania, and alumina particles with high sphericity, but are not limited to these.
The filler may be treated by well known techniques that render the filler hydrophobic with agents that react with the silylating agent and / or the adhesive matrix, for example using a coupling agent. Suitable coupling agents include silanes, titanates, and zirconates and aluminates. Suitable silylating agents include functional silylating agents, silazanes, silanols, and siloxanes, but functional silylating agents are not limited thereto. Suitable silazanes include hexamethyldisilazane (Huls H730) and hexamethylcyclotrisilazane, silylamides such as bis (trimethylsilyl) acetamide (Huls B2500), silylureas such as trimethylsilylurea, and silylimidazoles such as trimethylsilylimidazole. Is included, but is not limited to these.
Examples of titanate coupling agents include tetraalkyl type, monoalkoxy type, coordinated type, chelate type, quaternary salt type, neoalkoxy type, and cycloheteroatom type. Preferred titanates include tetraalkyl titanate, Tyzor ™ TOT (tetrakis (2-ethyl-hexyl) titanate), Tyzor ™ TPT (tetraisopropyl titanate), chelated titanate, Tyzor ™ GBA (titanium acetylacetylacetate). Nate), Tyzor ™ DC (titanium ethyl acetoacetonate), Tyzor ™ CLA (DuPont patent), monoalkoxy (Ken-React ™ KR TTS), Ken-React ™, KR-55 Tetra (2,2 diallyloxymethyl) butyl, di (ditridecyl) phosphite titanate, LICA ™ 38 neophenyl (diallyl) oxy, tri (dioctyl) pyro-phosphate titanate.
Suitable zirconates include any of the zirconates detailed on page 22 of the Kenrich catalog, in particular KZ55-tetra (2,2diallyloxymethyl) butyl, di (ditridecyl) phosphite zirconate, NZ-01- Neopentyl (diallyl) oxy, trineodecanoyl zirconate, NZ-09-neopentyl (diallyl) oxy, and tri (dodecyl) benzene-sulfonyl zirconate.
Aluminates that can be used in the present invention include diisobutyl (oleyl) acetoacetyl aluminate (KA301), diisopropyl (oleyl) acetoacetylaluminate (KA322) and KA489, although the aluminate is not limited to Kenrich ™. .
In addition to the above, the disclosed heat initially applied at a very high dilution (0.1-1.0% solution in MEK), for example vinyl benzene, vinyl pyridine, or the like, such as a cross-linked vinyl polymer. Certain polymers can be used, such as any size of curable matrix adhesive. Certain organic peroxides such as dicumyl peroxide can also be reacted with the filler.
The adhesive itself may be thermosetting or thermoplastic, such as polyglycidyl ether, polycyanurate, polyisocyanurate, bis-triazine resin, poly (bis-maleimide), norbornene-terminated polyimide, polynorbornene, acetylene. Terminal polyimide, polybutadiene and its functionalized copolymers, cyclic olefin polycyclobutene, polysiloxane, polyciscaloxane, functionalized polyphenylene ether, polyacrylate, novolac polymers and copolymers, fluoropolymers and copolymers, melamine polymers and copolymers , Poly (bisphenylcyclobutane), and blends or prepolymers thereof. It should be understood that the adhesives described above may be mixed together or mixed with other polymers or adhesives to add flame retardancy or increased toughness.
The average flow pore size and minimum pore size used here were determined using a Coulter ™ Porometer II (Coulter Electronics Ltd., Luton, UK) reporting the values directly. Average and maximum particle sizes were determined using a Microtrak ™ light scattering particle size analyzer, model number FRA (Microtrak Division of Lees & Northup, North Wales, PA, USA). Average particle size (APS) is defined as the value at which 50% of the particles are larger. Maximum particle size (LPS) is defined as the largest particle that can be detected with a Microtrak ™ histogram. Alternatively, the maximum particle size is defined as the smallest point when the Microtrak ™ model FRA Particle Analyzer indicates that 100% of the particles have passed.
In general, the method of adjusting the adhesive-fill dielectric is (a) sufficient to stretch the oil-sliding extruded preform to allow small particles and adhesive to flow freely into the gap or pore volume. Stretching and expanding polytetrafluoroethylene to a microstructure; (b) making a paste from a polymeric material, such as a thermosetting or thermoplastic material, and a filler; (c) dipping, coating, or The supply of pressure requires the adhesive-filler paste to be impregnated into a highly porous scaffold such as expanded polytetrafluoroethylene.
Table 1 shows the effect of the relationship between the mean flow pore size (MFPS) and particle size of the substrate. Insufficient results are observed when the ratio of mean flow pore size (MFPS) to maximum particle size is 1.4 or less. In this case, no uniform composite material is observed and most particulate filler does not penetrate uniformly into the microporous substrate. A uniform composite material is obtained when the ratio of MFPS to maximum particles is greater than about 2.0. It can also be seen that the higher the ratio of MFPS to largest particles, the more the microporous substrate will absorb the uniform dispersion.

Example 1
The fine dispersion was added to a 20% (w / w) MEK solution of polyglycidyl ether based on bisphenol-A (Nelco N-4002-5, Nelco Corp) catalyzed by flame retardant dicyanamide / 2-methylimidazole. 281.6 g TiO₂(TI Pure R-900, Du Pont Company) was mixed and adjusted. This dispersion was constantly stirred to ensure uniformity. Thereafter, a piece of expanded PTFE cloth was dipped in the resin mixture. The web was dried at 165 ° C. for 1 minute under tensile stress to give a flexible composite. A partially cured adhesive composite made in this way is 57% by weight TiO 2.₂13 wt% PTFE and 30 wt% epoxy adhesive. Several plies of adhesive sheet were stacked between the copper foils and pressed at 225 ° C. for 90 minutes at 600 psi using a vacuum assisted hydraulic press and then cooled under pressure. As a result, a copper having a dielectric constant of 19.0, an average layer thickness of 100 μm (0.0039 ″ (3.9 mil)) and a dielectric laminate thickness that can withstand 280 ° C. soldering impact for 30 seconds. The laminated board was made.
Example 2
The fine dispersion was 386 g SiO pretreated with phenyltrimethoxysilane (04330, Huls / Petrarch).₂(HW-11-89, Harbison Walker Corp.) was mixed with 200 g of bismaleimide triazine resin (BT206OBJ, Mitsubishi Gas Chemical) and 388 g of MEK in a manganese catalyst solution. This dispersion was constantly stirred to ensure uniformity. Thereafter, a 0.0002 ″ thick expanded PTFE fabric swatch was dipped into the resin mixture, removed, and then dried at 165 ° C. for 1 minute under tensile stress to give a flexible composite. Several plies of this prepreg were stacked between copper foils, pressed at 250 psi using a vacuum assisted hydraulic press at a temperature of 225 ° C. for 90 minutes, and then cooled under pressure, resulting in this The dielectric made is 53 wt% SiO₂It contains 5 wt.% PTFE and 42 wt.% Adhesive, adheres well to copper, has a dielectric constant of 3.3 (at 10 GHz) and a dielectric loss tangent (at 10 GHz) of 0.005. Indicated.
Example 3
The fine dispersion was 483 g SiO₂(HW-11-89) was prepared by mixing 274.7 g of bismaleimide triazine resin (BT2060BJ, Mitsubishi Gas Chemical) and 485 g of MEK with a manganese catalyst. This dispersion was constantly stirred to ensure uniformity. Thereafter, a 0.0002 ″ thick expanded PTFE fabric swatch was dipped into the resin mixture, removed, and then dried at 165 ° C. for 1 minute under tensile stress to give a flexible composite. Several plies of this prepreg were stacked between copper foils, pressed at 250 psi using a vacuum assisted hydraulic press at a temperature of 225 ° C. for 90 minutes, and then cooled under pressure, resulting in this The dielectric made is 57 wt% SiO₂It contains 4 wt% PTFE and 39 wt% adhesive, has good adhesion to copper and has a dielectric constant of 3.2 (at 10 GHz) and a dielectric loss tangent (at 10 GHz) of 0.005. Indicated.
Example 4
The fine dispersion was 15.44 kg TiO₂The powder (TI Pure R-900, DuPont Company) was prepared by mixing in 3.30 kg of bismaleimide triazine resin (BT206OBH, Mitsubishi Gas Chemical) and 15.38 kg of MEK manganese catalyst solution. This dispersion was constantly stirred to ensure uniformity. Then 0.0004 ”TiO₂Filled expanded PTFE fabric pieces (filled by the technique of Mortimer US Pat. No. 4,985,296, except TiO₂The addition amount was 40%, and the membrane was not compressed at the end), soaked in the resin mixture, removed and then dried at 165 ° C. for 1 minute under tensile stress to give a flexible composite. . The partially cured adhesive composite made in this way is 70% by weight TiO.₂9 wt% PTFE and 21 wt% adhesive. Several plies of this prepreg were stacked between copper foils and pressed at a temperature of 220 ° C. for 90 minutes at 500 psi using a vacuum assisted hydraulic press and then cooled under pressure. The dielectric thus obtained was in close contact with copper and showed a dielectric constant of 10.0 and a dielectric loss tangent of 0.008.
Example 5
The fine dispersion was 7.35 kg of SiO.₂(ADMATECHS SO-E2, Tatsumori) was prepared by mixing with 7.35 kg of MEK and 73.5 g of a coupling agent, ie 3-glycidyloxypropyltrimethoxysilane (Dynasylan GLYMO (Petrach System)). SO-E2 has a particle size of 0.4 to 0.6 mm and a specific surface area of 4 to 8 m.²/ G, and the manufacturer explains that it has very spherical silica with a bulk density of 0.2-0.4 g / cc (loose).
To this dispersion, 932 g of 50% (w / w) methyl ethyl ketone (MEK) solution of cyanated phenolic resin, Primaset PT-30 (Lonza Corp.), RSL 1462 (Shell Resins, Inc. (CAS # 25068-38- 6)) 50% (w / w) MEK solution 896 g, BC-58 (Great Lakes, Inc.) 50% (w / w) MEK solution 380 g, Bisphenol A (Aldrich Company) 50% MEK solution 54 g, Irganox 1010 (Ciba Geigy) 12.6 g, manganese 2-ethylhexanoate (Mn-HEX-CEM (OMG Ltd.)) 3.1% solution, and 2.40 kg of MEK were added. The dispersion was subjected to ultrasonic agitation through a Misonics continuous flow cell at a rate of about 1-3 gallons / minute for about 20 minutes. The finely dispersed system thus obtained was further diluted so that the solid content concentration of the entire bath was 11.9% (w / w).
The fine dispersion was poured into the impregnation bath. The expanded polytetrafluoroethylene web has the node fibril structure of FIG. 10 and has the following physical properties.

The Frazier number is related to the breathability of the substance being tested. Breathability is measured by securing the web to a gasketed gripping tool that is provided in a circular area of approximately 6 square inches for measuring air flow. The upstream side was connected to a flow meter in a pipeline with a source of dry compressed air. The downstream side of the sample gripper was released to the atmosphere. The test is performed by applying a 0.5 inch water column pressure upstream of the sample and recording the air flow through an in-line flow meter (ball float rotameter connected to the flow meter).
Ball burst strength is a test for measuring the relative strength of a sample by measuring the maximum value at the time of fracture. Test with a 1 inch diameter ball while the web is secured between the two plates. A Chatillon Force Gauge Ball / Brush Test was used. The medium is placed in tension on the measuring device and pressure is applied by lifting the web into contact with the ball of the burst probe. Record the pressure at failure.
The web described above was passed through a constantly stirring impregnation bath at a rate of about 3 feet / minute to ensure uniformity. The impregnated web was collected in rolls immediately through a heated oven to remove all or almost all of the solvent.
Several plies of this prepreg were stacked between copper foils and pressed at a temperature of 220 ° C. for 90 minutes using a vacuum assisted hydraulic press at 200 psi and then cooled under pressure. The dielectric thus obtained adhered well to copper and showed a dielectric constant (at 10 GHz) of 3.0 and a dielectric loss tangent (at 10 GHz) of 0.0085.
The physical properties of the particulate filler used in Examples 4 and 7 are compared below.

Example 6
SiO prepared from steam combustion of molten silicon₂An ePTFE matrix containing an impregnated adhesive filler mixture based on was prepared as follows. Two precursor mixtures were first prepared. One is in the form of a slurry in which the silica contains silane-treated silica as in Example 5, and the other is a catalyst-free blend of resin and other components.
Mixture I
This silica slurry is a 50/50 blend of SO-E2 silica in the MEK of Example 5, where the silica contains a coated silane equal to 1% of the weight of the silica. To a 5 gallon container, 17.5 pounds of MEK and 79 g of silane were added and the two components were mixed to ensure a uniform dispersion of the silane in MEK. Thereafter, 17.5 pounds of silica from Example 5 was added. The MEK-silica-silane mixture for two 5 gallon containers is added to the reaction vessel and the contents, ie, the slurry, are passed through an ultrasonic disperser for approximately 1 hour and recirculated to remove any silica aggregate that may be present. Destroyed. The contents were heated at a temperature of approximately 80 ° C. for approximately 1 hour while the sonication was complete, followed by continuous stirring of the reactor contents. The reacted mixture was then transferred to a 10 gallon container.
Mixture II
A given resin blend product is a MEK-based solution containing a catalyst-free resin blend (adhesive) containing approximately 60% solids, all solids in weight percent, PT-30 Exactly 41.2% cyanide phenolic resin, 39.5% RSL 1462 epoxy resin, 16.7% BC58 flame retardant, 1.5% Irganox 1010 stabilizer, and 1% bisphenol A promoter Mixture.
In a 10 gallon container, 14.8 pounds of PT-30 and 15-20 pounds of MEK were added and stirred vigorously to completely dissolve the PT-30. Then 6 pounds of BC58 was weighed and added to the MEK / PT-30 solution and stirred vigorously to dissolve the BC58. Stabilizer, 244.5 g Irganox 1010 and 163 g bisphenol A were added. A 10 gallon container was reweighed and 14.22 pounds of RSL 1462 was added. Additional MEK was added to weigh the mixture to 60 pounds. The contents were stirred vigorously for 1-2 hours or as long as the solid components were completely dissolved.
The desired product is a mixture of silica treated with silane, a resin formulation without catalyst, and MEK, with 68% solids by weight of silica and total solids by 5-50% by weight of the mixture. The exact solids concentration varies from experiment to experiment and depends in part on the membrane to be impregnated. The concentration of catalyst was 10 ppm of the sum of PT-30 and RSL 1462.
The solids content of Mixtures I and II were measured to test for precursor accuracy and to compensate for any solvent flushes that occurred so far. Mixture I was then added to a 10 gallon container to provide 12 pounds of solids, for example 23.48 pounds of Mixture I if 51.5% solids. Mixture II was then added to the vessel to provide 5.64 pounds of solids, for example 9.46 pounds of Mixture II if 59.6% solids. 3.45 g of manganese catalyst solution (0.6% in mineral spirits) was added to the mixture of Mixtures I and II and mixed thoroughly to make a high solids content mixture.
The mixture of baths to impregnate the ePTFE matrix was adjusted by adding enough MEK to this high solids content mixture where 28% was a solids mixture to bring the total weight to 63 pounds.
Thereafter, ePTFE matrix was impregnated with this bath mixture to make a dielectric material.
Example 7
The fine dispersion was mixed with 26.8 g of furnace black (Special Schwartz 100, Degussa Corp., Ridgefield Park, New Jersey) and 79 g of coupling agent, 3-glycidyloxypropyl-trimethoxysilane (Petrac System). It was adjusted. The dispersion was subjected to 1 minute of ultrasonic agitation followed by 17.5 pounds of SiO in 17.5 pounds of MEK pre-sonicated.₂(SO-E2) was added to the stirred dispersion. The final dispersion was heated at reflux with constant overhead mixing for 1 hour and then allowed to cool to room temperature.
Separately, the adhesive varnish was mixed with 57.5% (w / w) Primaset PT-30 MEK mixture 3413 g, 76.8% (w / w) RSL 1462 MEK mixture 2456 g, 53.2% (w / w). BC58 (Great lakes, Inc.) MEK solution 1495 g, 23.9% (w / w) Bisphenol A (Aldrich Company) MEK solution 200 g, 71.5 g Irganox 1010, 0.6% (w / w) Mu It was prepared by adding 3.21 g of a mineral spirit solution of HEX-CEM (OMG Ltd.) and 2.40 kg of MEK.
In a separate container, 3739 g of the above dispersion was placed with 0.0233 g of Furnace Black (Special Schwarz 100, Degussa Corp., Ridgefield Park, New Jersey), the adhesive varnish 1328, and 38.3 pounds of MEK. This mixture was poured into an impregnation bath and the ePTFE web was passed through the impregnation bath at a rate of 3 feet / minute or about 3 feet / minute. The dispersion was constantly stirred to ensure uniformity. The impregnated web was immediately passed through a heated oven to remove all or almost all of the solvent and collected on a roll.
Several plies of this prepreg were stacked between copper foils and pressed at 200 psi using a vacuum assisted hydraulic press at a temperature of 200 ° C. for 90 minutes and then cooled under pressure. The resulting dielectric showed good adhesion to copper.
Example 8
Adhesive varnish was mixed with 57.5% (w / w) Primaset PT-30 (PNP 88-1591) MEK solution 3413 g, 76.8% (w / w) RSL 1462 MEK solution 2456 g, 53.2% (w / w) BC58 (Great lakes, Inc.) MEK solution 1495 g, 23.9% (w / w) Bisphenol A (Aldrich Company) MEK solution 200 g, Irganox 1010 71.5 g, 0. It was adjusted by adding 3.21 g of a mineral spirit solution of 6% (w / w) Mn HEX-CEM and 2.40 kg of MEK.
In a separate container is 1328 g of the adhesive varnish dispersion, 42.3 lbs of MEK, 6.4 g of Furnace Black (Special Schwartz 100, Degussa Corp., Ridgefield Park, New Jersey) and SiO.₂(SO-E2) 1860.9 g was added. This mixture was poured into an impregnation bath and the ePTFE web was passed through the impregnation bath at a rate of 3 feet / minute or about 3 feet / minute. The dispersion was constantly stirred to ensure uniformity. The impregnated web was immediately passed through a heated oven to remove all or almost all of the solvent and collected on a roll.
Several plies of this prepreg were stacked between tin foils and pressed at 200 psi using a vacuum assisted hydraulic press at a temperature of 220 ° C. for 90 minutes and then cooled under pressure. The resulting dielectric showed good adhesion to copper.
As mentioned above, the present invention provides a dimensionally stable core for chip packages. This core is, for example, copper or a metal core such as INVAR plated with copper. Copper is one of the preferred materials because it makes it easier to match the CTE of the chip package to the CTE of the printed circuit board that bonds it. This makes the package easier to handle because the metal provides the necessary stability to the package. Plating the top and bottom surfaces of the chip package and laminating the layers simultaneously allows the chip package manufacturing process to be more efficient. In addition, the amount of metal in each layer on each side of the substrate is balanced to facilitate attachment of the package to the chip. Additional advantages of the metal core include providing embedded capacitance that reduces the simultaneous switching noise of the chip.
As shown in FIG. 11, a multilayer substrate using conductive Z-axis material may be used as taught in US Pat. No. 5,498,467, which is incorporated herein by reference. . For example, multiple layers can be bonded together using such conductive Z-axis material to obtain a very dense printed circuit board. One can think of reaching 100 or more layers. In FIG. 11, substrates 140 and 150 are each identical to each other and are structurally similar to the dimensionally stable substrate of FIG. Substrates 140 and 150 are laminated with a conductive Z-axis member 160 which will be described in detail below. FIG. 19 shows substrates 290 and 300 similar to the substrate of FIG. 8 that are laminated with a conductive Z-shaft 160 described below with reference to FIGS.
The multilayer package illustrated in FIG. 11 is described in US Pat. No. 5,498,467, cited above, with an irregularly shaped conduction path extending through the thickness of the Z-axis material. Such a conductive Z-axis material is used. The base material for forming the conductive Z-axis material is a flat, open-cell, porous member having continuous pores from one surface to the other. This porous, flat plate member has an internal configuration that creates an irregular path through which the material defining the pores passes in the Z-axis direction in a vertically defined cross-section through the Z-axis plane as shown in FIGS. Must have.
Suitable material for conductive Z-axis member is 5 × 10^-6m-5x10^-FourIt has a thickness in the range of about m (5 to 500 μm), and includes woven or non-woven fabric such as nylon, glass fiber or polyester fabric or cotton. This member may be a flexible porous polymer film or membrane, for example, a porous polyolefin such as porous polyethylene, porous polypropylene, porous fluoropolymer, or open cell, porous polyurethane. It may be such. Furthermore, open-cell, porous inorganic materials such as a thin porous ceramic plate having continuous pores connected from one surface to the other surface can be used.
Porous fluoropolymers include porous polytetrafluoroethylene (PTFE), porous expanded polytetrafluoroethylene (ePTFE), polytetrafluoroethylene and polyester or polystyrene porous copolymers, tetrafluoroethylene and fluorine. Ethylene-propylene (FEP) or C₁~ C_FourA perfluoroalkoxy-tetrafluoroethylene (PFA) copolymer having the following alkoxy group is included, but the porous fluoropolymer is not limited thereto. Preferred porous materials include expanded polypropylene, porous polyethylene and porous polytetrafluoroethylene. Most preferably, this material is expanded polytetrafluoroethylene having a microstructure in which the nodes are interconnected with fibrils and a porosity of 20-90%, for example, U.S. Pat. Such as a material prepared in accordance with the teaching of US Pat. No. 3,953,566.
Generally, the Z-axis member has a thickness of about 5 to 500 μm, preferably about 5 to 125 μm. However, the thickness is not a critical factor as long as the ultraviolet light penetrates the sample.
With reference to FIGS. 12-14, the material 170 for forming the Z-axis member is expanded polytetrafluoroethylene with pores 180 defined as spaces between nodes 190 interconnected by fibrils 200. The internal structure of the nodes 190 interconnected by the fibrils 200 results in an irregular continuous path through the Z-axis in the vertically defined cross-section 220 from one side 230 to the other side 240 of the plate member. A material density of the material. (See Figure 14)
As taught by US Pat. No. 5,498,467, a selectively conductive member 250 (FIG. 14) is provided by making the region 220 passing in the Z-axis direction susceptible to metal salt deposition. . This metal salt is converted to a non-conductive metal core upon irradiation energy, which then serves to catalyze the deposition of the conductive metal from the electroless metal plating bath. The pores 180 of the porous member 170 are initially moistened with a wetting agent such as alcohol or an organic aqueous surfactant. Methanol, propanol, tetrafluoroethylene / vinyl alcohol copolymer or the like can be used. The wetting agent serves to allow the material of the member to accept a conductive metal such as nickel or copper. Particularly preferred is copper.
The radiation-sensitive metal salt composition is a liquid radiation-sensitive composition that includes a solution of a photosensitive reducing agent, a metal salt, a halide ion source, and a second reducing agent. The radiation-sensitive solution contains water, a metal salt, a photosensitive reducing agent, a second reducing agent, and optionally a surfactant (if it is difficult to wet the surface). Metal salts include copper acetate, copper formate, copper bromide, copper sulfate, copper chloride, nickel chloride, nickel sulfide, nickel bromide, ferrous-containing compounds such as ferrous sulfate, ferrous chloride, and the like Including noble metals such as palladium, platinum, silver, gold, and rhodium, but the containing salts are not limited to these.
Suitable photosensitive reducing agents include aromatic diazo compounds, iron salts such as ferric or ferrous oxalate, ammonium ferric sulfate, etc., dichromates such as ammonium dichromate, anthraquinone disulfonate or salts thereof. , Glycine (especially active in wet surface conditions), L-ascorbic acid, azide, and other similar, and metal promoters, for example, tin compounds such as stannous chloride, or silver , Palladium, gold, mercury, cobalt, nickel, zinc, iron and the like, the latter group being optionally added in an amount of 1 mg to 2 g per liter.
The second reducing agent may be a polyvalent such as glycerol, ethylene glycol, pentaerythritol, mesoerythritol, 1,3-propanediol, sorbitol, mannitol, propylene glycol, 1,2-butanediol, piconal, sucrose, dextrin. Compounds such as alcohol and triethanolamine, propylene oxide, polyethylene glycol, lactose, starch, ethylene oxide and gelatin are included, but the second reducing agent is not limited to these. Compounds that can also be used as the second reducing agent include aldehydes such as formaldehyde, benzaldehyde, acetaldehyde, and n-butyraldehyde, and triamides such as nylon, albumin, and gelatin, and 4-dimethylaminotriphenylmethane, 4 ′, 4 Leuco bases of triphenylmethane dyes such as' 4 ''-tri-di-methylamino-triphenylmethane, 3,6-bisdimethylaminoxanthene and 3,6-bisdimethylamino-9- (2-carboxy It includes leuco bases of xanthene dyes such as ethyl) xanthene, polyethers such as ethylene glycol diethyl ether, diethylene glycol, diethyl ether tetraethylene glycol dimethyl ether, and other similar ones, but they are used as secondary reducing agents. Compounds which can be is not limited thereto.
The second reducing agent, which is also a humectant exemplified by sorbitol, is used in one embodiment as a moisturizing component of the processing solution, because of its moisturizing effect on the film that is clearly “dried” before development. The It substantially assists in maintaining the metal coating density of the internal material of the member during the development process in which all unconverted radiation sensitive composition in the coating is washed away with a base.
Suitable surfactants include Rohm & Hass Co. There are non-ionic surfactants based on the reaction of nonylphenol and glycidol, such as polyethenoxy nonionic ethers such as TRITON X-100 manufactured by Olympus, and surfactants 6G and 10G from Olin Mathieson Company.
This treatment solution, i.e. the radiation-sensitive composition, usually comprises an acidifying agent in the form of an acid salt and a small amount to adjust the pH of the aqueous solution to between 2 and 4 (preferably 2.5 to 3.8). Of halide ions (iodine, bromine or chlorine ions), so that the combination of additives substantially increases the density of the subsequently formed coating by exposing the treated plate material to irradiation energy Give a good effect. Adjusting acidity does not necessarily require the introduction of a drug only for that purpose. Because the adjustment can be performed with other functions, as illustrated by photosensitive reducing agents of acidic nature (eg, ascorbic acid, glycerin, etc.) or some other additive (eg hydrochloric acid) to introduce halide ions. This is because it may be completely or partially made by an acidic substance having a large amount. Similarly, some or all halide ions may be introduced as a component of a metal salt that can be reduced (eg, cupric).
Among the many suitable acidic substances that can be used to control or adjust the pH of the sensitized solution are fluoroboric acid, citric acid, lactic acid, phosphoric acid, sulfuric acid, acetic acid, formic acid, boric acid, hydrochloric acid, nitric acid and others There are similar ones. A wide variety of bromide, chloride and iodide salts and other halogen-generating compounds may be utilized to provide some or all of the required halide ion content of the processing solution. These include, among others, generally salts of metals with these halogens, cupric bromide, nickel chloride, cobalt chloride, cupric salt, sodium iodide, potassium iodide, lithium chloride, magnesium iodide, bromide. Such as illustrated by magnesium, sodium bromide, potassium bromide, and the like can be included. Bromide salts are more preferred because they result in higher sensitivity on the substrate than the corresponding chlorides (ie, denser and denser precipitates) at least in certain cases.
Halide ions constitute only a smaller percentage of the solute, typically about 0.045 to 1.6%, preferably about 0.13 to 0.45%, based on the total weight of dissolved solids. The range is acceptable. In other terms, the amount of halogen is about 0.9 to 25 milliequivalents, preferably 2.5 to 9 milliequivalents per liter of sensitive solution, such as 0.3 to It may be 1.0 g / liter. Increasing the proportion of halide ions is usually not preferred because such an increase will gradually reduce the susceptibility effect of the treatment to less than that obtained in an optimal amount. Also, the proportion of these halide ions expressed as equivalents is less than that of cupric or other reducing non-noble metal cations in the treatment solution. For example, the ratio of equivalents of such metal ions to halide ion equivalents is usually in the range of at least 2: 1, preferably 4: 1 to 100: 1.
The radiation sensitive composition is applied to the material 170 to completely wet the material defining the pores, thereby allowing the porous member to pass through or penetrate the pores 180 of the material 170 so that the porous material is porous. The planar material is exposed to the radiation sensitive composition for a time sufficient to form a coating within the pores along the material defining the pores from one side to the other. Thereafter, the coated porous member is naturally dried or dried by heating in a furnace at 50 ° C. or lower. At this stage, the material should be processed under yellow light conditions to preserve the photosensitivity of the processing composition. Also, the member should be kept at a temperature below 70 ° F. and a relative humidity of 60% or less because of the potential for water absorption by the member material which can adversely affect processing.
A selected location on one surface of the coated porous member is masked with an opaque covering 260 (FIG. 12) to prevent subsequent irradiation from hitting the covered area. The mask 260 results in any desired shape, size, arrangement of punctiform conductive regions, or conductive regions separated by every other band of nonconductive regions and passing through in the Z-axis direction. Alternate bands or strips can be produced. The points are usually circular, but could be other geometric shapes such as squares, rectangles. The dot size can be as small as 0.0001 inch and as large as 0.025 inch, preferably 0.001 inch, 0.002 inch, 0.003 inch, 0.004 inch, 0.005 inch, 0 .008 inches, 0.009 inches, or fractions thereof, where the pitch defined as the distance between the centers of adjacent points is preferably at least twice the size of the points, for example 1 mil In the point, the pitch is 2 mil.
Enough to reduce the masked member 170 to radiation, such as light, electron beam, x-rays, and the like, preferably ultraviolet light, to reduce the metal cations of the metal salt through the thickness of the member into metal nuclei. Exposure with sufficient time and sufficient power. After that, the mask of the member 170 is removed and washed with an acidic or alkaline cleaning solution to wash away the radiation-sensitive composition protected by the opaque cover. If the acidic or alkaline cleaning (or fixing) solution is not left in contact for longer than a few minutes, for example 5 minutes or less, the solution will not affect the areas where the radiation has reduced metal cations to metal nuclei.
Specifically, as shown in FIG. 12, the member 170 to be processed is selectively masked with a metal mask, diazo or silver halide film. The masked member 170 is then photographed with an uncollimated or collimated ultraviolet light source with a wavelength shorter than 500 nm. In order to create a stable photo image, the catalyst, the non-conductive metal core, itself requires a minimum of 200 mJ of irradiation energy.
This ultraviolet light energy is strong enough to penetrate the thickness of the porous member. In this way, in the subsequent plating process, the conductive metal is continuously plated through the Z axis to give electrical continuity to the Z axis. If necessary, ultraviolet light energy can be applied to both sides of the planar member.
After a 5 minute normalization period, the catalyzed material is treated with a sulfuric acid solution for a short period of 30-90 seconds, for example, with a solution of 8 wt% sulfuric acid and 92 wt% deionized water, or hydroxylated Wash with an alkaline solution of 40 g / l ethylenediaminetetraacetic acid, 100 ml / l formaldehyde, adjusted to a pH higher than 10 with sodium. The purpose of this washing step is to remove the unexposed catalyst from the material while retaining the photoreduced image.
The washed material containing the selective image is then stabilized with a reactive metal cation displacement solution.

The image undergoes a substitution reaction of copper with a more stable cation, such as palladium. Such a thin layer amount tends to oxidize copper faster, and palladium is also capable of initiating a reduction reaction in an electroless solution faster, so a more stable system is desired. The member is left in this solution for at least 30 seconds and then washed in deionized water for about 1 minute.
The catalyzed member is selectively electrolessly plated with one or more conductive metals to a deposition thickness of about 50-60 microinches. Such metals include copper, nickel, gold, silver, platinum, cobalt, palladium, rhodium, aluminum and chromium. While being placed in the electrolytic bath, the member is swung to agitate and promote the diffusion of metal to the deepest region of the substrate. The plating is first rinsed in deionized water and then immersed in a stirred electroless copper bath for a sufficient time to deposit copper on the palladium in the material and through the thickness of the substrate. Done.
In this manner, within selected regions that penetrate the material in the Z-axis direction, the material nodes 190 and fibrils 200 are formed with a conductive metal layer 230 having a Z-axis portion 290 and upper and lower contact pads 270 and 280, Cover at least partially. The conductive metal forms a continuous conductive path 290 through selected areas between the upper and lower pads 270 and 280.
In the following examples, the catalyst treatment solution used was prepared by adding the following to 1 liter of deionized water.

The fixing solution used was 8% by weight sulfuric acid and 92% by weight distilled water. A stabilizer is also used and contains the following ingredients:

Example 9
As taught by US Pat. No. 5,498,467, W.W. L. By immersing the stretched porous polytetrafluoroethylene membrane obtained from Gore & Associates in a solution of 75% methanol, 25% ethanol and 1% by weight copolymer of tetrafluoroethylene and vinyl alcohol at room temperature for about 30 minutes. And treated with a wetting agent.
The wet membrane was then immersed in the catalyst treatment solution for 60 seconds and then dried in a 50 ° C. oven for 3 minutes. Thereafter, one surface of the membrane was masked with a diazo film having points of 6 mil diameter and 6 mil pitch (center to center).
The membrane was then exposed to a 1600 mJ collimated ultraviolet light source for about 2 minutes. After a 5 minute break-in period, the UV treated film was washed in the fixer solution for 30 seconds to remove the unexposed catalyst treatment solution. Next, the selectively drawn membrane was stabilized by immersing in the stabilizing solution for 1 minute, and then washed with distilled water for 1 minute.
The stabilized membrane is then based on per liter of deionized water,

Was immersed in a copper plating bath composition (Shipleys 3).
The membrane was stirred in this bath using a stir bar for 7.5 minutes to facilitate the diffusion of copper throughout the pores of the catalyzed portion of the membrane through the Z axis.
The conductive Z-axis material path can be filled with an adhesive to act as a connector interface between two other conductive materials. Suitable adhesives include epoxy resins, acrylic resins, urethane resins, silicone resins, polyimide resins, cyanate ester resins, and the like. Usually, the adhesive is absorbed in the pores by immersing the film in a solution of the adhesive. In the case of epoxy resins, a suitable solvent is methyl ethyl ketone. Once the adhesive has been absorbed or impregnated into the conductive Z-axis material, it is fired at 160 ° C. to provide a B-stage conductive Z-axis member in order to provide bonding capability.
Example 10
A stretched, porous polytetrafluoroethylene membrane, 6 mil (150 μm) thick, was wetted by immersion in 2 propanol for 1 minute. It was then immersed in the catalyst treatment solution for 1 minute as in Example 9, dried, masked and exposed to ultraviolet light. It was then treated with the fixing and stabilizing solution as in Example 9 and then plated with copper as in Example 9.
Example 11
A 2 mil thick stretched porous polytetrafluoroethylene membrane was prepared as in Example 9 except that the masking strip was a 3 mil pad with a 10 mil pitch.
Examples 12-14
The procedure of Example 9 for a membrane made from porous polyethylene, porous polypropylene and continuous pore porous polyurethane to make a conductive Z-axis material with an irregularly shaped conductive path extending along the Z-axis. Followed.
Example 15
A film made from a stretched porous polytetrafluoroethylene film having a node-fibril structure shown in FIG. 15 (1000 times magnification) has a density of 0.22 g / cm at 25 ° C.^ThreeAnd an air volume of 70% and a thickness of 76 μm. L. Available from Gore & Associates and adjusted as in Example 9 except that the strip to be masked to make a conductive Z-shaft was a 2 mil pad with a 5 mil pitch.
Example 16
W. L. Density 0.4 g / cm at 25 ° C, available from Gore & Associates^ThreeA stretched porous polytetrafluoroethylene membrane with a node-fibril structure shown in FIG. 16 (1500 times magnification) with a thickness of 40 μm and an air volume of 20% is a stripping mask to make a conductive Z-axis membrane Was adjusted as in Example 9 except that the 8 mil pad was a 15 mil pitch.
Example 17
Density 0.35g / cm at 25 ° C^ThreeAnd an expanded porous polytetrafluoroethylene membrane having a node fibril structure of FIG. L. To make a conductive Z-axis material, available from Gore & Associates, it was adjusted as in Example 9 except that the strip to be masked was an 8 mil pad with a 15 mil pitch.
Example 18
Density 0.20g / cm at 25 ° C^ThreeAnd an expanded porous polytetrafluoroethylene membrane having a node fibril structure of FIG. L. To make a conductive Z-axis material available from Gore & Associates, it was adjusted as in Example 1 except that the masking strip was an 8 mil pad with a 15 mil pitch.
The conductive Z-axis material of Examples 10-18, including the conductive paths described above, can be filled with an adhesive to provide a connector interface between two other conductive materials. Suitable adhesives include epoxy resins, acrylic resins, urethane resins, silicone resins, polyimide resins, cyanate ester resins, or the like. Usually, the adhesive is absorbed into the pores by immersing the film in an adhesive solution. For epoxy resins, a suitable solvent is methyl ethyl ketone. As in Example 9, the Z-axis material to provide bonding capability is impregnated with an epoxy adhesive and fired at 160 ° C. to provide a B-stage Z-axis composite material.
Although specific embodiments of the invention have been illustrated and described herein, the invention is not limited to that shown and described. It will be apparent that changes and modifications may be incorporated and embodied as part of the present invention within the scope of the appended claims.

Claims (17)

(A) a metal core having at least one clearance made therethrough;
(B) Dielectric layers disposed on the upper surface and the lower surface of the metal core, respectively, wherein the dielectric layer is a stretched polytetrasilane holding a mixture containing a granular filler and an adhesive resin A ratio of the average flow pore size of the expanded polytetrafluoroethylene material to the average particle size of the particulate filler is greater than about 1.4, and the mixture includes the expanded expanded polytetrafluoroethylene material. Distributed substantially uniformly throughout the pore volume of the tetrafluoroethylene material;
(C) a metal cap layer disposed on each of the dielectric layers; and (D) a conductive via electrically connecting the metal cap layers;
A chip package substrate. (A) a metal core having at least one clearance made therethrough;
(B) Dielectric layers disposed on the upper surface and the lower surface of the metal core, respectively, wherein the dielectric layer is a stretched polytetrasilane holding a mixture containing a granular filler and an adhesive resin A particulate fluoroethylene material, wherein the mixture is substantially uniformly distributed throughout the pore volume of the expanded polytetrafluoroethylene material and has an average flow pore size of the expanded polytetrafluoroethylene material. The ratio of filler to maximum particle size is at least about 2;
(C) a metal cap layer disposed on each of the dielectric layers; and (D) a conductive via electrically connecting the metal cap layers;
A chip package substrate. (A) a metal core having at least one clearance made therethrough;
(B) Dielectric layers disposed on the upper surface and the lower surface of the metal core, respectively, wherein the dielectric layer is a stretched polytetrasilane holding a mixture containing a granular filler and an adhesive resin The particulate filler, wherein the mixture is substantially uniformly distributed across the pore volume of the expanded polytetrafluoroethylene material, and is the minimum pore size of the expanded polytetrafluoroethylene material. The ratio of to the maximum particle size is at least about 1.4;
(C) a metal cap layer disposed on each of the dielectric layers; and (D) a conductive via electrically connecting the metal cap layers;
A chip package substrate. The chip package substrate according to claim 1, wherein the metal core is copper. The chip package substrate according to claim 1, wherein the metal core has a thickness of about 1.4 mil (35.6 μm) or less. The chip package substrate according to claim 1, wherein the dielectric layer has a thickness of about 5 μm or more. 4. The chip package substrate of claim 1, 2 or 3, wherein the dielectric layer has a thickness in the range of about 25 [mu] m to about 50 [mu] m. The chip package substrate according to claim 1, wherein the adhesive resin is a cyanate ester resin. 4. The chip package substrate according to claim 1, wherein the metal core is a nickel alloy containing 30 wt% to 42 wt% nickel having a copper sheet laminated on an upper surface and a lower surface thereof. (A) a step of making a metal core having a clearance;
(B) disposing a first dielectric layer on the upper surface of the metal core and disposing a second dielectric layer on the lower surface of the metal core, wherein the first and second dielectric layers; Each having an expanded polytetrafluoroethylene material holding a mixture containing a particulate filler and an adhesive resin, wherein the mixture is substantially uniform across the pore volume of the expanded polytetrafluoroethylene material. The ratio of the average flow pore size of the expanded polytetrafluoroethylene material to the average particle size of the particulate filler is greater than about 1.4 ;
(C) disposing a metal cap layer on each of the upper surface of the first dielectric layer and the lower surface of the second dielectric layer;
(D) laminating first and second dielectric layers and a metal cap layer on each surface of the metal core; and
(E) forming at least one conductive via that electrically connects the metal cap layers on each side of the metal core through the clearance of the metal core;
A method for manufacturing a chip package substrate, comprising: (F) A step of repeating steps (b) to (e) 2 to 100 times;
The method of claim 10 , further comprising: The laminating step (d)
(I) applying a pressure of approximately 300 to 350 psi (2.07 to 2.41 MPa) to the dielectric layer and the metal cap layer;
(Ii) heating at a rate of temperature increase of 5-7 ° C. per minute until a temperature of 177 ° C. is reached;
(Iii) maintaining a temperature of 177 ° C. for approximately 30 minutes;
(Iv) raising the temperature to 220 ° C. to 225 ° C. and holding at this temperature for approximately 60 minutes; and (v) slowly cooling the chip package while maintaining the pressure;
The method of claim 10 , comprising: The method of claim 10, wherein the metal core is copper. The method of claim 10, wherein the metal cap layer is a copper cap layer. The method of claim 10, wherein the first and second dielectric layers are thermosetting prepregs. The method of claim 15 , wherein the thermosetting prepreg is a nonwoven material comprising a cyanate ester resin in a polytetrafluoroethylene matrix. The method of claim 16 , wherein the first and second dielectric layers have a thickness sufficient to fully satisfy the clearance of the metal core.

Images